See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/366322032

Smart grid evolution: Predictive control of distributed energy resources—A

review

Article  in  International Journal of Electrical Power & Energy Systems · December 2022

DOI: 10.1016/j.ijepes.2022.108812

CITATIONS READS
2 416

5 authors, including:

Oluleke Babayomi Zhenbin Zhang

Shandong University Shandong University
42 PUBLICATIONS   143 CITATIONS    223 PUBLICATIONS   3,078 CITATIONS   

SEE PROFILE SEE PROFILE

Tomislav Dragicevic Jiefeng Hu

Technical University of Denmark Federation University Australia
401 PUBLICATIONS   15,582 CITATIONS    165 PUBLICATIONS   3,685 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Smart microgrids View project

Fault-Tolerant Cascaded H-bridge Multilevel Inverter for Medium-Voltage Variable Speed Electric Motor Drive View project

All content following this page was uploaded by Oluleke Babayomi on 19 December 2022.

The user has requested enhancement of the downloaded file.

This article has been accepted for publication in International Journal of Electrical Power & Energy Systems. This is the author's version which has not
been fully edited and content may change prior to final publication. Citation information: DOI 10.1016/j.ijepes.2022.108812
1

Smart Grid Evolution: Predictive Control of

Distributed Energy Resources—A Review
Oluleke Babayomi, Zhenbin Zhang, Tomislav Dragicevic, Jiefeng Hu, and Jose Rodriguez

Abstract—As the smart grid evolves, it requires increasing techniques. Among several candidate methods, model predic-
distributed intelligence, optimization and control. Model pre- tive control (MPC) has been gaining increasing popularity
dictive control (MPC) facilitates these functionalities for smart among researchers who work on different aspects of the
grid applications, namely: microgrids, smart buildings, ancil-
lary services, industrial drives, electric vehicle charging, and smart grid. MPC facilitates the superior performance of the
distributed generation. Among these, this article focuses on multivariable constrained optimization of several applications,
providing a comprehensive review of the applications of MPC namely: microgrids [3], [4], smart buildings [5]–[7], grid
to the power electronic interfaces of distributed energy resources ancillary services [8], industrial automation [9], [10], plug-
(DERs) for grid integration. In particular, the predictive control in electric vehicles (PEV) [11], [12], and grid-connected DER
of power converters for wind energy conversion systems, solar
photovoltaics, fuel cells and energy storage systems are covered power electronics converters [13].
in detail. The predictive control methods for grid-connected con- Microgrids are intelligent and automated modular grids
verters, artificial intelligence-based predictive control, open issues which operate autonomously, and can be connected to the
and future trends are also reviewed. The study highlights the po- main grid too [14]. They comprise DERs, distribution feeder
tential of MPC to facilitate the high-performance, optimal power lines and diverse kinds of loads. The application of MPC to
extraction and control of diverse sustainable grid-connected
DERs. Furthermore, the study brings detailed structure to the AC and DC microgrids facilitates multivariable and multi-
artificial intelligence techniques that are beneficial to enhance time-scale implementation in primary, secondary and tertiary
performance, ease deployment and reduce computational burden hierarchies [15]. Primary control via MPC attends to volt-
of predictive control for power converters. age/current, frequency, power sharing, virtual impedance, ESS
Index Terms—Smart grid, distributed energy resources, model management, and power quality control. Secondary control
predictive control, power electronic converter, microgrid, dis- regulates the compensation of voltage/frequency deviations,
tributed generation, grid-connected converter, artificial intelli- optimal power flow and black start management [16]–[19].
gence
The tertiary level attends to market participation [20], multi-
microgrid coordination [21], [22] and optimal power dispatch
I. I NTRODUCTION [22], [23]. Details on these topics can be found in related
A smart grid is an electricity grid with bidirectional review articles [3], [4].
power/data flow [1], and integrated advanced informa- Smart building MPC applications are implemented for de-
tion/communication, sensing, measurement and control tech- mand response, i.e., price-based and incentive-based programs
nologies. These features facilitate the smart grid’s flexible, [24]. Higher electricity prices are associated with peak-demand
reliable, resilient, stable, and sustainable operation [1], [2]. period, and incentivization is an added program for users
The concept of a modern grid was motivated by the need who reduce loading during peak hours. It operates by the
to [2]: 1) improve efficiency of electricity production and DSO broadcasting an emergency signal (with energy price and
distribution, 2) improve reliability, 3) empower electricity duration of scheduled load shedding) to all clients for load
users with information to control their electrical power usage shedding or shifting. At this point, smart controllers (using
and costs, and 4) mitigate the climatic impacts of the electrical convex optimization or other optimization techniques) at the
power industry. These led to industrial, research and regulatory user’s end effect the request on electrical devices. Demand
actions for the evolving smart grid. response problems can be solved with MPC formulated as
Four elements enable the smart grid to deliver the afore- convex optimization problems (for improved stability and
described functionalities, viz., distributed energy resources robustness) [24]. MPC can provide optimal allocation of power
(DERs), information communication technologies and sensors, references for building energy flexibility (from energy storage,
vehicle-to-grid infrastructure, and electricity markets. Fig. 1 heat pump with thermal storage, plug-in EVs), and load
shows the electrical infrastructure at the generation, trans- power consumption [25]. Furthermore, additional objectives
mission, distribution and consumption stages of the grid. In including pricing and CO2 intensity (of hybrid energy sources)
the smart grid, these elements interact by the bidirectional can be included in the cost functions [26].
dataflow of control signals and measurement data from sensors Ancillary services are specialty services which facilitate
and smart meters over secure information and communication reliable power supply in the grid. They are supplied by
channels. Internet of things (IoT) facilitates the cyber-physical specialty providers to the system operator [27]. The most
monitoring and control of smart grid elements (see Fig. 1). essential are frequency and voltage control services. Frequency
The large number of control variables for the diverse smart control involves maintaining the frequency at regulatory levels
grid elements and timescales of control require optimal control by ensuring balance between active power generated and
 2

GENERATION TRANSMISSION DISTRIBUTION CONSUMPTION

P, Q

MMC
Offshore wind park
DC line
MMC
Distribution system
Electricity
P, Q
HVDC Transmission P, Q
operators (DSO) market
Solar farm DC/DC Prosumers with

microgeneration
Battery storage DC  Microgrid AC  Microgrid
DC line
H2 MMC Load Load PCC P, Q
Hydrogen fuel cell DC/DC PEV Charging

+ V2G feature
P, Q

HVAC Transmission
Thermal plants
Load Load H2

AC  Microgrid DC  Microgrid

Hydropower P, Q Smart building
Multi-microgrid network 

Sensors/actuators Sensors/actuators Sensors/actuators Sensors/actuators

Generation controllers Transmission controllers Distribution controllers User controllers

` `

Cyber layer

Fig. 1. Overview of the smart grid showing the cyber-physical layers.

consumed [27]. The deployment of positive and negative The optimal scheduling of plug-in electric vehicle (PEV)
frequency control reserves helps to achieve this goal. Voltage charging constitutes a multi-objective problem with con-
control service regulates participating devices which generate straints, and can be solved with predictive control. The
or absorb reactive power as a means to control voltage levels objectives are to charge numerous PEVs connected to the
[28]–[30]. As the smart grid accommodates an increasing distribution network, while maintaining the bus voltage level
number of converter-interfaced DERs, MPC becomes bene- within the regulatory limits. These are considered in different
ficial to implement optimal frequency [31], [32] and voltage studies which carry out combined charging scheduling and
control [33], [34]. The predictive control of a large number power control [39], plug-and-play [40], and demand response
of heterogeneous thermostatically controlled loads to provide participation [41]. The cyber-physical configurations of PEV
ancillary service was also validated in [35]. It was shown that charging stations can be grouped into three: centralized, hier-
by engaging ancillary services within an MPC framework, the archical [42], and decentralized [11], [12]. Centralized control
microgrid operating expenses can reduce by almost 25% [8]. configuration has a central control hub where all optimization
MPC has been also demonstrated for the performance control for the operation of the numerous charging stations
enhancement of high power, medium voltage drives used in is carried out subject to several constraints including grid
industrial machines. In particular, it has faster and more robust constraints and user constraints. This method faces severe
responses than linear cascaded control schemes like field- limitations as the scale of charging points increases. Therefore,
oriented control [36]. Nonetheless, with high power drives, hierarchical centralized control provides a solution to the
lower power converter switching frequencies are necessary challenge of scale by reducing both processing hardware
to reduce the switching losses. Hence, multi-step (or long) and communication requirements [11]. This involves grouping
prediction horizons are favored [37], [38]. It was reported the charging stations into local charging units, area charging
that an algorithm with comparable performance to the multi- centers and wide area charging centers. The decentralized con-
step methods, namely, MPC with optimized pulse patterns, has figuration operates with a more democratic control structure;
been commercially deployed by ABB [10]. each charging unit has the ability to communicate intelligently
 3

(through is onboard controller) with the grid and dictate how regulate maximum power-point tracking (MPPT) for optimal
it wants to operate. Customer preference can be prioritized power extraction from the DER. The grid-end converter MPC
and communicated to the distributed system operator. Also, regulates optimal power exchange with the grid. Bidirectional
the control scheme is more robust to communication channel energy flow is necessary for energy storage charging and
disturbances than centralized and hierarchical control methods. discharging cycles (wind turbines with regenerative capability
The smart grid has several options for the sustainable also cause bidirectional flow). The focus systems in this study
distributed generation of electrical energy from DER such as are the power converters directly connected to the DER, and
solar photovoltaic (PV), wind, fuel cell, ocean, tidal, wave, the grid-connected converter, as highlighted in Fig. 2.
bioenergy, and energy storage systems (ESS). These DERs
usually need to be operated at their operating points of maxi- A. Introduction to Model Predictive Control
mum power extraction. Hence, MPC enables high performance
Model predictive control (MPC) is a nonlinear control tech-
control of power electronic conversion for solar photovoltaic
nique which emerged in the process industry in the 1970s [10].
(PV) systems [43], wind energy conversion systems (WECS)
It has emerged as a popular control technique with application
[44], fuel cells [45] and energy storage systems (ESS) [13].
in cross-disciplinary domains. This is due to its high dynamic
Authors in [46] studied MPC applications for PV systems
performance, and capability for constrained optimization [3].
only, and did not provide further information for other re-
Over the past half-century, it has found industrial relevance
newable sources. The review of optimal energy management
to petrochemical, aerospace and automotive processing and
in microgrids [47], and MPC for microgrids in primary, sec-
manufacturing.
ondary and tertiary hierarchical levels were reported in [3], [4].
MPC is an optimization-based control technique that utilizes
However, only microgrid applications were covered. The use
a dynamic process model to predict the future evolution of the
of MPC in optimal energy management was reviewed by [48];
system’s state and output. The control objectives are formu-
but it only applies to the tertiary control of energy sources.
lated as optimization problem with the system inputs being
The review of MPC for power converters in electric drives
the optimization variables. At each time step, this problem
was reported in [36]; but it mainly applies to industrial drives.
is solved over the prediction horizon whereby the process
A recent review covered MPC’s applications to microgrid
model yields the effect of the input sequence on the objective
DERs [49]. This study investigated MPC’s high performance
function. Finally, the first input value is applied to the real
control of PV, wind and energy storage systems. Nonetheless,
system. MPC has the inherent ability to handle multiple-input
the study excludes several mathematical details, and does not
and multiple-output control problems and takes constraints
include grid-connected converters and artificial intelligence-
regarding actuators, states, and outputs into consideration. For
based MPC for DERs.
all MPC algorithms, there are common elements, namely: the
Motivated by the literature gap, the objective of this article prediction model, and objective function (or cost function in
is to provide a comprehensive review on the state-of-the-art of the case of minimization problem). These will be discussed in
MPC applications to power electronic converter interfaces for the following.
DERs’ grid integration. The scope of this study covers the fol-
lowing related fields: i) renewable energy (and energy storage)
technologies [50], [51], and their integration in modern power B. System Prediction Model
systems [52]; ii) new smart transmission grid technologies The generic discrete model of a physical system or process
for ancillary services support in power systems [53]–[55]; requiring regulation is given by (1), where k ∈ N is the
iii) smart distributed and autonomous energy systems [56], discrete time step, x is the state vector (with state variables
including AC and DC microgrids [3], [4], [57], and DERs [58]. x ∈ x), y is the output vector (with output variables y ∈ y).
The rest of the article is organized as follows: The introduction
x(k + 1) = f (x(k), u(k)), (1a)
to MPC is discussed in section II. The applications of MPC to
converter-interfaced DERs is covered in section III. The pre- y = x(k). (1b)
dictive control of grid-connected DER-converters is covered
in section IV. Artificial intelligence techniques for predictive Given a sequence of input variables over a prediction hori-
control are discussed in section V. MPC-based control of zon Np ∈ N time steps. We can define a sequence of possible
virtual power plants and grid ancillary services is discussed control input states that the controller could implement as [10],
in section VI. Open issues and future trends are discussed in [59]:
section VII. Finally, the conclusion is drawn in section VIII.
U(k) = [uT (k) uT (k + 1) ... uT (k + Np − 1)]. (2)
II. P RELIMINARIES ON MPC FOR DER S YSTEMS The predictive controller seeks an optimal control input se-
quence
Fig. 2 depicts a typical DER (comprising wind, solar PV,
fuel cells and battery energy storage (BESS)) and interfacing Uopt (k) = min J,
U(k)
systems which facilitate its connection to the grid. The stages (3)
of the system include primary energy source and storage, s.t. x(k) ∈ X , u(k) ∈ U
the interfacing power converters (back-to-back DER-side and where J is the cost function that captures the control objectives
grid-side), and grid-connected filter. At the DER, MPC can for the optimization problem, X and U are the state and input
 4

Focus system in Sec. III Focus system in Sec. IV

Grid-side converter
DER DER-side conv.
Wind turbine,
ESS Vdc

Solar PV,

fuel cells Filter Grid

S=P+jQ
Real-time predictive controllers

Fig. 2. Power electronic conversion system for grid-connected DER.

constraints sets respectively. The control objective would be to Fig. 4 (a) depicts the MPC-based regulation of modular multi-
track a reference y∗ (k + 1) by minimizing the tracking error level power converters (MMC) for high voltage DC (HVDC)
magnitude as [10] transmission of electrical energy from offshore wind farms to
k+Np −1
the grid. MMC inner structure is depicted in Fig. 4 (b).
X In the following subsections, the theory and application of
J= kyp (k + 1) − y∗ (k + 1)k2Q + λu k∆u(k)k22 , (4)
MPC to power electronic converters for WECS, ESS, fuel
k=1
cells and solar PV will be covered in detail. Table I presents
where Q is the penalty matrix on the tracking error, yp comparative details of MPC applications to different types of
is the predicted output which is derived from the system’s DER converters.
mathematical model, y∗ is the reference, λu penalizes the
control effort ∆u(k), and ∆u(k) := u(k) − u(k − 1). MPC
is a receding horizon scheme; therefore, only the first term of A. Predictive Control of Generator-Side WECS Converters
the sequence Uopt (k) is applied to the plant. Recalculation of In this subsection, applications of MPC to DER-side power
the next sampling instant’s predictions and optimal sequence converters for wind energy conversion systems (WECS) will
is done after updated estimates of the states. Fig. 3 depicts be discussed in detail.
how MPC is applied to a typical plant in the smart grid. Electricity generated from wind turbines in 2020 was 1270
TWh, representing 3.7-fold increase from 2010 [111]. This
Optimization problem Control input
is because between 2010 to 2020, the cost of generating
min. (3) sequences electricity from onshore wind, and offshore wind sytems
fell by 56% and 48% respectively [112]. Wind generators
 
Constraints are commercially made as squirrel-cage induction generator
Cost function (4)

(SCIG), wound rotor induction generator (WRIG), doubly-fed
Plant
induction generator (DFIG), permanent magnet synchronous
State prediction (1) generator (PMSG) and wound rotor synchronous generator
(WRSG) [113]. Among these, PMSG and DFIG are commonly
Sensor/
studied in the literature, and will be further discussed in this
Disturbances Prediction model
estimator article. This study focuses on the application of predictive
control to WECS; details on WECS hardware configurations
can be found in [113].
Fig. 3. Model predictive control of a plant in the smart grid [5]. Plant
represents different DERs, grid-connected filters and electrical loads. Fixed-speed (±1%) generators are the oldest form of
WECS, and they have limitations which include [113]: (i)
lower energy conversion efficiency, (ii) wind speed variability
impacts grid frequency directly, (iii) mechanical stresses on
III. P REDICTIVE C ONTROL OF D ISTRIBUTED E NERGY
the WECS mechanical components during grid faults. On the
R ESOURCES
other hand, variable-speed WECS (especially those with full-
The smart grid has several options for the sustainable dis- variable speeds of 0 − 100%), overcome these challenges by
tributed generation of electrical energy from DER such as solar (i) higher conversion efficiency, (ii) decoupling the wind speed
photovoltaic (PV), wind, fuel cell, ocean, tidal, wave, bioen- variations from grid frequency, (iii) reducing wear and tear
ergy, and energy storage systems (ESS). These DERs need of mechanical components (e.g., a PMSG can be directly
to be operated at their operating points of maximum power coupled without a gearbox), (iv) improving power quality,
extraction. Hence, MPC enables optimal high performance and (v) reducing acoustic noise [114]. Nonetheless, because
control of power electronic conversion for solar photovoltaic they require power electronic converters, they cost more than
(PV) systems [43], wind energy conversion systems (WECS) their fixed-speed counterparts. The generator-side converter—
[44], fuel cells [45] and energy storage systems (ESS) [13]. voltage source rectifier (VSR)—facilitates the operation of the
 control
DC-link controller

with MPPT

Real-time controller 5

(a) DERs with grid-connected back-to-back converters

Offshore generation Wind-farm MMC HVDC

Grid-tied MMC
transmission

DC line

AC Grid 1

Wind farms

DC line

AC Grid 2

Model predictive

Model predictive

current control
Real-time
controller power control
with MPPT

(a) DER with HVDC transmission to the grid

SM1
+
+
_
_
SM2

Upper arm

SMn
+
+
_
_

(i) Single-phase equivalent circuit

SM1

Lower arm

SM2

+
Vc +
_
VSM _
SMn

(ii) AC equivalent circuit (iii) DC equivalent circuit

(b) Structure of the grid-connected MMC

Fig. 4. Predictive control of grid-connected DER with HVDC transmission.

wind-turbine generator at variable wind speeds/frequencies, a) MPC for Multi-level Boost Converter for PMSG
therefore, extending the efficiency of wind energy conversion. WECS: The MPPT algorithm maximizes the reference speed
to the PMSG for variable wind speed.
1) Permanent Magnet Synchronous Generator: PMSGs Control objectives and cost function: Let us consider a
have their rotor excitation supplied by permanent magnets, generic voltage source converter with NL voltage levels (where
saving about 30% of generator losses (arising from DC excita- NL is a postive integer greater than or equal to 2). The
tion) [115]. Thus, they have high power density and efficiency, control objectives for the direct model predictive power control
with low lifetime costs. PMSGs used for direct-driven WECS (DMPPC) of the NL -level converter include: (i) the tracking
are usually smooth-surface multi-pole machines, also having of the power reference; (ii) balancing of the DC-link capacitor
neglible salience [116]. There are two common converter voltages (in the case when NL > 2), and (iii) minimization of
topologies on the WECS generator-side [117]: i) voltage the control input (switching frequency). The DMPPC scheme
source active-front-end rectifier (with maximum power-point will be regulated by the cost function Jp in Table II, where
tracking (MPPT) control), and ii) passive rectifier and DC/DC each term represents objectives (i)-(iii) respectively. Further
boost converter (with MPPT control). details on the model can be found in [118].
 6

TABLE I
A PPLICATIONS OF PREDICTIVE CONTROL TECHNIQUES TO DER CONVERTERS

DER Type Control Method Feature Application

Wind energy Single-vector XOne voltage vector per sampling period. XPMSG [60]–[65]
conversion system predictive control XHigher ripple content. XDFIG [66], [67]
(WECS)
Multiple-vector predic- XTwo or more voltage vectors per sampling period. XPMSG [68], [69]
tive control XLower ripples and tracking errors. XDFIG [70]–[72]

Computationally XReduces computational efforts. XPMSG [64], [73]–[76]

efficient DMPC XUtilizes hexagon or triangle candidate region, dis- XDFIG [72], [77], [78]
crete space vector modulation (SVM) etc.
XConstant switching frequency with SVM
Solar PV FCS-MPC-MPPT XImproved conversion efficiency, fast dynamic re- XBuck converter [79]
sponse, than linear control. XFlyback converter [80]–[84]
XNegligible oscillations around the maximum XBoost converter [85]
power point.
CCS-MPC-MPPT XMore grid-friendly frequency spectrum than FCS- X Boost converter [86], [87]
MPC-MPPT

DO-MPC-MPPT XPerforms better than FCS-MPC-MPPT under XFlyback converter [88], [89]
rapidly changing weather conditions. XBoost converter [90]
Energy storage FCS-MPC XFaster control dynamics than linear control. XMultilevel flying-capacitor converter
system (ESS) XOptimized transient performance within system [91]–[93]
constraints. XBuck converter [94], [95]
XBuck-boost [96], [97]
XDual active bridge converter [98]
XBoost converter [99]
XPulsed power loads [100], [101]
XConstant power loads [102], [103]
Fuel cell MPC XIncreaseses life span of the fuel cell. PEMFC [104], [105]
XMaximizes the active catalytic surface area.
Hybrid DER MPC XOptimal energy management and power sharing XPV-wind-battery [106]
management for hybrid energy storage system. XPV-wind-hydrogen fuel cell [107],
[108]
XPV-ESS [109], [110]
FCS-MPC is finite control set MPC, CCS-MPC is continuous control set MPC, DO-MPC is discrete observer MC, MPPT is maximum power point
tracking, PMSG is permanent magnet synchrous generator, and PV is photovoltaic.

b) Direct Model Predictive Torque Control (DMPTC) of ripples. Thus, the dual-vectors DMPTC scheme gives a better
PMSG WECS: First, the classical DMPTC is discussed, and performance.
two variants are introduced afterwards. Dual-vectors DMPTC: This is an improvement over the
Control objectives and cost function: Let us consider a classical DMPTC through the application of two voltage
generic rotor-side voltage source converter (VSC) with NL vectors within a sampling interval, which has two variants. The
voltage levels (where NL is a postive integer greater than or first utilizes an active and a zero vector [119]. Through duty
equal to 2). The control objectives for the DMPTC of the cycle optimization, a fraction of control period is designated
NL -level converter are (i) tracking of the torque reference, for an active (non-zero) voltage vector and the remaining time
(ii) tracking of PMSG current reference, (iii) maintenance of for a zero vector. This method faces severe limitations in
maximum torque Temax constraint, and (iv) balancing of the that the resultant voltage vector always lies in phase with the
DC-link capacitor voltages (in the case when NL > 2). The active vector, resulting in sub-optimal phase [120]. The second
DMPTC scheme will be regulated by cost function JDMPTC variant utilizes two active voltage vectors, producing optimal
in Table II where each term represents objectives (i)-(iv) length and phase, and with lower ripples [121], [122]. The
respectively. latter selects the optimal pair from any neighboring active
Classical DMPTC: Consider the case of a two-level con- vectors. Let the chosen optimal active vectors be {uq , ur }.
verter; DMPTC will evaluate the cost function for the finite Their optimal application times will be [121], [122]
set of applicable voltage vectors ui ∈ U := {u0 , u1 , . . . , u7 }.
A single optimal voltage vector uopt is applied over the entire ∂J(im (k + 1), Te (k + 1), (uq , ur ))
=0 (6)
sampling duration Ts . ∂tq , ∂tr
s.t. (i) tq + tr = Ts , (ii) tq , tr ∈ [0, Ts ],
uopt := arg min J(ui )(im (k + 1), Te (k + 1)) (5)

Nonetheless, this method has limitations because the chosen

vector could be distant from the ideal (which may lie in- ∂J(im (k + 1), Te (k + 1), (uq , ur ))
between the finite voltage vectors), resulting in relatively high = 0, (7)
∂tq , ∂tr
 7

Cost
Phase
Length

Time

function
optimization
optimization

Compensation
min. (10) (15),(16) (11)

Gate

signal

generation (a) Multiple-vector direct predictive torque control

Cost
Length

Time

function
optimization

Compensation
min. (10) (11)

Gate

signal

generation (b) Single-vector direct predictive torque control

Fig. 5. Direct predictive torque control techniques for PMSG WECS [116].

s.t. (i) tq +∂tr = Ts , (ii) tq , tr ∈ [0, Ts ], where tx is the optimal scheme will be regulated by cost function JMPDPC in Table
time to apply ux ∀ x ∈ {q, r}. After pairs of (ux , tx ) have II, where each term represents objectives (i)-(iii) respectively.
been enumerated, the applied pair is derived by optimization. The conventional MPDPC predicts active and reactive
power from the discretized form of the derivative of negative
(uq , tq , ur , tr ) := arg min J(ui )(im (k + 1), Te (k + 1)). (8) d
complex apparent power (− dt S). Nonetheless, this procedure
Multiple-vectors DMPTC: This method applies an addi- is not efficient in the selection of optimal voltage vectors
tional step to the dual-vector DMPTC. After utilizing active and increases computational resources. Hence, a seminal low-
vectors to obtain (uq , ur ), if they do not satisfy system complexity MPDPC (LC-MPDPC) introduced by [123] will
constraints, they are are combind with a voltage vector uz . be described.
First, (uq , ur ) are synthesized to become uopt [68], [116], Low-Complexity MPDPC (LC-MPDPC): MPDPC selects
[119]: an optimal voltage vector that minimizes the power error
tq tr
uopt = uq + ur . (9) [124], and produces better accuracy and lower power ripples
Ts Ts than linear control e.g., direct power control with look-up
Hence, the optimal times for (uopt , uz ), i.e., (topt , tz ), respec- table [125], [126]. Nonetheless, classical MPDPC requires
tively, are computed by [119] a rigorous evaluation of all candidate voltage vectors in the
∂J(im (k + 1), Te (k + 1), (uopt , uz )) optimization procedure, and may require several prediction
=0 (10) horizons for accuracy. Thus, [123] proposed an efective LC-
∂topt , ∂tz
MPDPC within a single prediction horizon.
s.t. (i) topt + tz = Ts , (ii) topt , tz ∈ [0, Ts ], (11)
LC-MPDPC appplies the principles of direct current control
and (iii) uz ∈ {u0 , u7 }. [127] to select the most optimal voltage vector within one
Thus, if n =
topt
∈ [0, 1], then prediction horizon. LC-MPDPC reduces the procedure in
Ts
conventional MPDPC from computations for eight voltage
topt opt
q = ntq , tr = ntr , and tz = Ts − (topt opt
q + tr ). (12) vectors (case of two-level converter) to only two vectors—an
active vector and a zero vector. The following steps describe
The afore-discussed variants of DMPTC are shown in Fig. 5.
the MPDPC algorithm:
2) Doubly-Fed Induction Generator (DFIG): DFIGs have
both rotor and stator windings connected to the external 1) Calculate S with instantaneous power theory [128].
three-phase AC terminal. The rotor winding is designed to 2) Predict the complex power S0 (k + 1) that is due to
operate flexibly with variable wind speeds, and this can supply zero vector u(000). The predicted power is the dis-
variable-frequency bidirectional active and reactive power to cretized derivative of the negative complex apparent
d
the grid though interfacing back-to-back converters. The stator power (− dt S).
supplies power at grid frequency [114]. 3) Obtain the error due to the zero vector, i.e., Serror
0 (k +
a) Model Predictive Direct Direct Power Control 1) = −(S∗ − S0 (k + 1)), where (.) ¯ is the complex
(MPDPC) of DFIG WECS: First, the classical MPDPC is dis- conjugate operator.
cussed, and variants of the improved low-computation MPDPC 4) Compute the actual angle of Serror
0 (k + 1), i.e., ∠Serror
0 ,
(LC-MPDPC) are introduced afterwards. by adding the angle of the grid voltage ∠eg . (It is
Control objectives and cost function: Let us consider a assumed that steps 1 to 3 were done in the d−q reference
generic voltage source converter with NL voltage levels (where frame.) Based on ∠Serror
0 , determine the active voltage
NL is a postive integer greater than or equal to 2). The control vector uopt closest to Serror
0 (k + 1).
objectives for the MPDPC of the NL -level converter include: 5) Compute the error vector Serror u (k + 1) =
(i) the tracking of the active power reference; (ii) the tracking −(S∗ − Su (k + 1)), due to uopt .
of the reactive power references, (iii) balancing of the DC-link 6) Compare Serror
0 (k + 1) and Serror
u (k + 1); the optimal
capacitor voltages (in the case when NL > 2). The MPDPC voltage vector between them has the smallest angle. E.g.
 8

Gear
Gear
Gear

box DFIG
box DFIG
box DFIG

DC link
DC link
DC link

Variable speed
Variable speed
Variable speed

and frequency
Vdc
and frequency
Vdc
and frequency
Vdc

AC-DC AC-DC AC-DC

Min. error angle    Duty
2nd active vector
min. optimization
+ duty optim.

Min. error angle  

State prediction Min. error angle  
abc abc
abc
(a) (b) (c)

(a) Classical Model Predictive Direct Direct Power Control  (c) Multiple-vector Low-Complexity MPDPC (LC-MPDPC):
Fig. 6. Predictive direct power control techniques (b)
for Single-vector
DFIG WECS. (a) Classical MPDPC
Low-Complexity model predictive direct
(LC-MPDPC): power control (MPDPC), (b) Single-vector low-
complexity MPDPC, (c) Multiple-vector low-complexity MPDPC.

the optimal vector is u0 (000) if ∠Serror

0 < ∠Serror
u <, 1) Smart Grid ESS Technologies and Converters: ESS
and uopt otherwise. consist of several mature and developing technologies which
Dual- and Multiple-Vector LC-MPDPC: The dual-vectors operate on electro-chemical, electro-thermal and electro-
LC-MPDPC improves the steady-state performance reported mechanical principles. Batteries energy storage and super-
in [123] by applying the duty-cycle optimization concept [129] capacitors are electro-chemical devices; flywheels and com-
to select optimal active and zero vectors. In order to achieve pressed air energy storage (CAES) are electro-mechanical; and
error-free steady-state performance, the multiple-vector LC- thermal energy storage systems are electro-thermal in nature
MPDPC which applies three optimal vectors (two adjacent [132], [133]. Fig. 7(a) shows that the ESS with higher power
active and one zero vector) was reported by authors in [130]. density have shorter discharge rate of seconds, while ESS with
Nonetheless the high-fidelity of steady-state performance was higher energy densities discharge over minutes to several days.
at the expense of higher switching losses. A control scheme Fig. 7(b) shows how these energy technologies are matched to
which applies four voltage vectors per sampling period was the parts of the grid: ESS with lower power rating and lower
reported in [70]. This method improves the other MPDPC discharge time are applied by end users and in the distribution
schemes by effective operation under both balanced and unbal- network; ESS with higher rating and higher discharge time are
anced conditions, and constant switching frequency. However, applied for transmission, generation and system management.
it is more complex to implement. The afore-discussed variants In order to maximize the diversity of electro-chemical
of MPDPC are illustrated in Fig. 6. characteristics of ESS types, hybrid ESS are common. A
Performance Comparison: Using switching losses (which supervisory controller regulated by MPC can assign optimal
are directly proportional to average switching frequency [131]) power references to the DC/DC controller of each storage
and power ripples, the comparative performances of the above type based on state of charge, power/energy density matching
techniques can be quantitatively assessed. Ref. [130] showed [134]–[136]. Solutions for hybrid ESS including battery [137],
that power ripples decrease as the applied voltage vectors [138], ultra/super-capacitors [139], [140] and flywheel [141]
per control period increases. However, there is a loss-trade- have been reported. In addition, optimal hybrid ESS-based
off because the switching frequency (hence, switching losses) frequency response for grid-support facilitates extended ESS
increases as the number applied vectors per control period lifetime and lower life-cycle costs [142]. A battery (with
increases. lower power density) will only be required to provide energy
for small changes in active power (and frequency), while
the ultracapacitor (with large power density) regulates large
changes in active power [142].
B. Predictive Control of Bidirectional DC/DC Converters for
ESS require bidirectional power flow through DC/DC con-
ESS
verters for charging and discharging. These converters are
This section covers the predictive control of power con- grouped as isolated and non-isolated; the isolated converters
verters for energy storage systems (ESS). It starts with a have galvanic isolation to physically decouple the input circuit
description of microgrid ESS technologies and converters. from the output, while non-isolated ones have physical conti-
Next it discusses MPC for bidirectional DC/DC converters for nuity between the input and output. The common non-isolated
ESS. Finally, MPC solutions to constant power load (CPL) DC-DC converters include half-bridge, inverting buck-boost,
and pulsed power load challenges with DC/DC converters are cascaded buck-boost, Cuk, SEPIC/Zeta, and switched capac-
discussed. itor converters [143], [144]. MPC was applied to interleaved
 9

Distribution

Days

Days
End User Generation Management
Pumped hydro & Transmission
Demand-side 
T&D asset
Electricity

management
invest. deferral
markets

Compressed air Demand

Flow bat. Congestion

Backup storage
Demand

Hours

response

Hours
management
balancing

NaS Thermal markets

Discharge Rate
Discharge Rate

Pb-acid1
Li-ion Batteries Peak load

Capacity firming
Capacity firming

smoothing

Minutes

Minutes
Pb-acid Bulk ancillary

services

NiCd
Black-start

NiMH
Frequency

Flywheel Power quality

regulation

Seconds

Seconds
Voltage control
Power quality

Supercapacitors

1kW 100kW 1MW 100MW 1GW 1kW 100kW 100MW 1GW

ESS Power Rating ESS Power Rating

(a) (b)

Fig. 7. Energy storage technologies and their functionalities in the smart grid. (a) Energy storage technologies and their application contexts. (b) Functionalities
of ESS in the smart grid. Acronyms: Flow battery (flow bat.), Sodium sulphide battery (NaS), Advanced lead acid battery (Pb-acid1 ), Lead acid battery (Pb-
acid), Lithium ion battery (Li-ion), Nickel Cadmium battery (NiCd), Nickel-metal hydride battery (NiMH), Thermal energy storage (Thermal), Compressed
air energy storage (Compressed air), Pumped hydro storage (Pumped hydro).

converters with robust control of uncertainties [145], [146], C. Predictive Control of Converters for Fuel Cells
and matrix converter for renewable generation [147]. Common Fuel cells have a high potential in the renewable energy
isolated converters have a high frequency transformer, and transition due to their ability to convert hydrogen gas to
include dual-active bridge (DAB), and flyback converters. electricity with very little greenhouse gases emissions (water
Although the afore-mentioned converters are suitable for low- being the main by-product). Also, they have an efficiency of
voltage applications, in medium to high voltage applications, 40 − 60%, making them up to three times more efficient than
DAB are preferred for their higher efficiency and relatively solar PV and two times more efficient than wind turbines
lower cost than modular multi-level DC/DC converters [148]. [157]. What’s more, they have almost ten times more energy
MPC-based DAB regulation with superior dynamics was density than batteries [158]. They are classified into six types
reported in [149]–[151]. Table I shows the applications of based on the type of fuel and electrolyte: (i) Proton exchange
MPC to bidrectional DC/DC converters for battery energy membrane fuel cell (PEMFC), (ii) alkaline fuel cell, (iii)
storage, and supercapacitor [95]. A predictive control scheme phosphoric acid fuel cell, (iv) molten carbonate fuel cell (v)
for a buck-boost bidirectional DC/DC converter for ESS has solid oxide fuel cell, and (vi) direct methanol fuel cell.
the cost function JESS (see Table II) for regulating battery Fuel cells are electronically conditioned by unidirectional
charging/discharging. DC/DC power converters to regulate their electrical power
outflow [159]. As earlier mentioned, non-isolated converters
are common. However, when galvanic isolation and higher
2) Predictive Control Solutions to Constant Power Load
voltage conversion rations are needed, isolated converters with
and Pulse Power Load Issues: A constant power load (CPL) is
integrated high-frequency transformers are utilized.
a tightly regulated electronic load that absorbs constant power,
Studies on MPC applications to fuel cells are reported in
and so, manifests negative impedance characteristics [152].
[45], [104], [105], [160]. In the maximum power/efficiency
This results in reduced damping and instability challenges. A
tracking operation of a PEMC, artificial neural network (ANN)
composite offset-free continuous MPC solution was proposed
was used to predict states of optimally controlled variables.
for a buck converter in [153]. Explicit MPC was applied to
MPC can also prolong device lifespan: The optimal power
a boost converter, with fuzzy-control of nonlinearities [154].
tracking performance of a PEMFC, without sacrificing the
Stability of a DC-bus via ESS injection current was studied
longevity of the cell stacks, can be achieved by multi-
in [102], [103].
objectively controlling cathode and anode pressures, and hy-
drogen/oxygen supply to the cells [104]. The cost function to
Pulsed power loads are associated with applications that optimally track the reference current is JF C in Table II.
draw high power within a brief moment as in electric ve-
hicles, electric ships and electric aircraft [101]. To prevent
instability, ESS with high power density, e.g., supercapacitors D. Predictive Control of Solar PV Converters
and flywheels, are necessary [100]. MPC techniques, including Electricity generated from solar PV in 2020 was 577 TWh,
continuous MPC [155], and explicit MPC [156], which have representing 18.9-fold increase from 2010 [111]. This can be
fast dynamic responses are effective for pulsed power control attributed to 85% drop in costs of utility-scale PV systems
of DC/DC boost converter. between 2010 to 2020 [112]. Relative to other renewable
 10

systems, PV cells have a low conversion efficiency. Hence, sampling time Ts , the state predictions become [46]
maximum power tracking (MPPT) controllers are usually
required to extract maximum power from them at all operating Ts
iL (k + 1) =iL (k) + (vPV (k) − (1 − Sn )vdc (k)),
conditions, especially during low insolation [161]. The com- L
mon MPPT algorithms include the perturb and observe [162], Ts
vdc (k + 1) =vdc (k) + (iPV (k) − (1 − Sn )idc (k)),
and incremental conductance [163] techniques. However, their C
most common limitations are: (i) non-convergence to the true vPV (k + 1) =(1 − D)vdc (k + 1), (14)
maximum power point during rapidly changing atmospheric
conditions, and (ii) higher oscillations around the maximum where D is the duty ratio.
power point [46].
JP V = λi (i∗L − iL (k + 1))2 + λv (vPV
∗
− vPV (k + 1))2 , (15)

References i∗L and vPV

∗
are determined by MPPT alogorithms
(perturb and observer or incremental conductance).
2) Digital-Observer
Fuel

MPC MPPT: The preceding method

calculates PV voltage and current references with conventional
Cell

Boost converter
MPPT algorithms. However, these are non-robust. Hence,
Boost converter
PV Panels
the digital-observer MPPT was introduced [89] for improved
robustness under dynamic atmospheric conditions. It utilizes
Prediction
Prediction

model

Min.
present and historical measurements to model the PV source
Min.

Cost 
model

as an equivalent voltage veq (22)
Req in (17) [89].
and resistance Cost 

function

These are used to compute references for thefunction

converter control.
MPPT (23)

Ref. Calc.
vPV (k) − vPV (k − 1)
FCS-MPC-MPPT control of DC/DC boost converter for PV system
Req = − (16)
Fig. 8. FCS-MPC-MPPT control of DC/DC boost converter for PV system. i (k) − i (k − 1)
Predictive control
PVof DC/DC boost
PVconverter for fuel cell conditioning

MPC solutions have been applied to address non- vPV (k) − vPV (k − 1)
convergence and instability issues of linear MPPT algorithms Req (k) = − ,
iPV (k) − iPV (k − 1)
for boost converter [161], buck converter [79], and buck- veq (k) =vPV (k) + Req iPV (k). (17)
boost converter [90]. MPC-based MPPT methods are grouped
into continuous control set-MPC-MPPT (CCS-MPC-MPPT) So, the predicted operating point tracks the maximum power
[86], [87] and discrete-MPC-MPPT (DMPC-MPPT). DMPC- point on the PV I-V curve by the cost function J
MPPT (18).
MPPT comprises two subgoups: finite control set MPC-
MPPT (FCS-MPC-MPPT) Battery [79]–[85], and digital observer-
Buck-boost converter vPV (k + 1) =vPV (k) ± ∆vPV (k),
MPC-MPPT (DO-MPC-MPPT) [88]–[90]. MPC-based meth-
ods introduce improved conversion efficiency, fast dynamic iPV (k + 1) =(veq (k) − vPV (k + 1))/Req ,
response, and negligible oscillations around the maximum
Prediction
Min.
P PV (k + 1) =vP V (k + 1)iP V (k + 1),
model

power point. DO-MPC-MPPT generally has better perfor- Cost 

JMPPT =PPV (k + 1) − PPV (k), (18)

mance than FCS-MPC-MPPT under rapidly changingfunction
weather
(21)

Ref. Calc.
conditions (because FCS-MPC-MPPT relies on conventional ∆vPV is an adaptive step-size which ensures that the maxi-
MPPT calculation methods). An improved method to reduce mum power point is closely tracked, and JMPPT is the cost
Predictive
drift of DO-MPC-MPPT control of
under bidirectional
highly DC/DC buck-boost
unsteady converterwas
conditions for battery energy storage.
function.
reported in [79]. The afore-discussed DERs can also be combined in hybrid
1) Classical MPC: The cost function JP V (15) in Table format. For instance, MPC can be applied to provide optimal
II is minimized by the classical MPC. This also relies on references to control hybrid combinations of PV-wind-battery
conventional MPPT techniques, and these reduce its overall [106], PV-wind-hydrogen fuel cell [107], and PV-ESS [109].
dynamic performance.

IV. P REDICTIVE C ONTROL OF G RID -C ONNECTED

d 1 C ONVERTERS
iL = (vPV − (1 − Sn )vdc ),
dt L
d 1 In this section, the predictive control of grid-connected
vdc = (iPV (1 − Sn ) − idc ), (13) converters will be discussed. First, the recent regulatory
dt C
control requirements for grid-connected converters will be
highlighted. Second, the recent research advances which meet
where Sn = 0 if S1 is OFF, and Sn = 1 if S1 is ON. those requirements through fixed switching frequency MPC
Discretizing (13) by the forward-Euler approximation, with solutions will be covered.
 11

TABLE II
C OST FUNCTIONS FOR POWER CONVERTERS IN DISTRIBUTED ENERGY RESOURCES

Application Equation Description

∗ (k + 1)

2
Boost Converter Jp = Pdc − Pdc (k + 1) Pdc (k + 1) := vdc (k + 1)idc (kP + 1) is the extrapolated
Permanent Magnet +λvc (vo (k + 1))2 + λsw u2sw (k + 1) reference power, usw (k + 1) = x=1,2,3 |ux (k + 1) −
Synchronous Generator ux (k)|, vdc , idc are predicted boosted converter input voltage
(PMSG) WECS and current respectively, and the voltage difference between
the upper and lower dc link capacitors vo (k + 1) = vc1 (k +
1) − vc2 (k + 1).

DMPTC PMSG WECS JDMPTC = λTe (Te∗ − Te (k + 1))2 vo (k + 1) is the predicted voltage difference between multi-
+λi (i∗g − i∗g (k + 1))2 + λmax (Temax − Te (k + 1)) level DC-link capacitors (vo (k + 1) is set to constant zero if
+λv (vo (k + 1))2 NL = 2).

Dual-fed induction gen- JMPDPC = λP (Ps∗ − Ps (k + 1))2 (x)∗ is the reference value of x ∀ x ∈ {Ps , Qs }; the stator
erator (DFIG) WECS +λQ (Q∗s − Qs (k + 1))2 + λv (vo (k + 1))2 active and reactive power are defined by Ps = Re(S),
Qs = Im(S), respectively; vo (k + 1) is the predicted voltage
difference between multi-level DC-link capacitors (vo (k + 1)
is set to constant zero if NL = 2).

Energy Storage JESS = (i∗b − ib (k + 1))2 i∗b is the reference for battery current ib , and SOC represents
s.t. SOCmin ≤ SOC(k) ≤ SOCmax , ib ≤ |imax | state of charge.

Fuel cell JF C = (i∗L − iL (k + 1))2 iL is the filter inductor current with reference i∗L .

Solar PV JP V = λi (i∗L −iL (k+1))2 +λv (vPV

∗ −v
PV (k+1))
2 iL is the converter filter inductor current with reference i∗L .
(vPV is the instantanous PV voltage with reference (vPV ∗ .
References i∗L and vPV
∗ are determined by MPPT alogorithms
(perturb and observer or incremental conductance).
WECS = wind energy conversion system, DMPTC = direct model predictive torque control.

A. Control Requirements Since CCS-MPC schemes have their fixed switching fre-
Finite control set MPC (FCS-MPC) is the more popular quency guaranteed by the modulator, the literature focuses
type of predictive control applied to DER converters. This is on innovating the optimal control of objectives for VSC,
mainly due to the absence of a modulator, which improves its modular multilevel converters (MMC), neutral-point clamped
speed of dynamic response. One recurring challenge with FCS- (NPC) converters, cascaded H-bridge, and inductor-inductor-
MPC is that it has variable switching frequency, which results capacitor (LLC) resonant converters for solar PV systems. The
in non-determistic harmonic spectra. This poses a challenge VSC applications feature embedded integrator and Kalman
for grid-connected converters which must meet regulatory har- filter [166], full and reduced-order models with closed-form
monic specifications, e.g., as specified by IEEE 519 Standard expressions [167]. Authors in [168] report on virtual voltage
(see Fig. 9). Thus, for a short circuit ratio ksc < 20, the vectors with space-vector PWM. A modulation-based MPC
maximum current harmonic at the point of common coupling technique was reported in [169] for normal and unbalanced
(iPCCn ) to the grid, for any harmonic order, is 4.0%. This grid conditions. Renewable systems like solar PV require
implies that conventional FCS-MPC will be unsuitable for MPPT control, thus fixed-switching frequency predictive phase
grid-connected systems if the harmonic components are not shift MPPT [170] and MPPT with space vector modulation
regulated. Several studies report solutions in this regard, and [170] are beneficial.
they will be discussed below. It should be clarified that among FCS-MPC comprises two variants: optimal switching vector
them, only [164] confirms that both fixed swithing frequency MPC (OSV-MPC), and optimal switching sequence MPC
and harmonic spectrum identical to PWM or SVM (with lower (OSS-MPC). The former being more commonly reported
switching frequency) are guaranteed. It achieves this desirable in the literature. As earlier mentioned, FCS-MPC has an
result by specifically controlling the switching symmetry in inherent variable switching frequency—in particular, OSV-
each sampling period. MPC. Therefore, a little more intricacy is necessary to achieve
fixed switching frequency. The literature reports the following
B. Solutions with Fixed Switching Frequency successful implementations with constant switching frequency:
Fixed switching frequency in MPC for grid-connected multiple vector FCS-MPC [172], floating virtual voltage vec-
power converters can be realized with continuous control- tors [171] and virtual vectors in a reshaped and compacted
set MPC (CCS-MPC), and finite control-set MPC (FCS- solution space [173]. Detailed methods and features of the
MPC). CCS-MPC requires a separate modulator apart from above categories are presented in Table III.
the optimizer. On the other hand, FSC-MPC combines both A predictive control technique based on floating virtual
the optimization and modulation within the same scheme, voltage-vector is shown in Fig. 10 [171]. The algorithm divides
requiring no separate modulator. the total vector area into 30 virtual vectors, instead of the 8
 12

TABLE III
MPC FOR GRID - CONNECTED CONVERTERS WITH FIXED SWITCHING FREQUENCY

Category Application Feature

Continuous VSC with L(CL) filter XEmbedded integrator and Kalman filter [166].
control-set MPC XFull and reduced-order models with closed-form expressions [167].
(CCS-MPC) XVirtual voltage vectors and space-vector PWM [168].
XMinimizes AC line and circulating current ripples [174].
XEffective for unbalanced grid [169].
Modular multilevel converter XOptimal modulation references [175].
XSliding-discrete-control-set modulated MPC [176].
3L neutral-point-clamped converter XAdvanced switching sequences [177].
XModulated MPC for grid current and dc-link capacitor voltages [178].
XMPPT and space vector modulation [170].
Cascaded H-bridge converter XSeparation of voltage balancing control from the cost function [179].
XModulated integral action MPC [180].
Switched-boost common-ground 5L inverter XSingle-step MPC [169].
LLC resonant converter for PV XFixed frequency predictive phase shift MPPT technique [170].
Finite control-set VSC with L(CL) filter XMultiple vector FCS-MPC [172].
MPC (FCS-MPC) XFloating virtual voltage vectors [171].
XVirtual vectors in a reshaped and compacted solution space [173].
XFixed modulation cycle similar to discontinuous pulsewidth modula-
tion [164].
H-bridge neutral-point-clamped converter XEliminates high-frequency common-mode voltage components.
[181].
XWeighting factor only affects the peak current during transients [182].
3L neutral-point clamped converter XCascaded optimal switching sequence MPC without weighting factor
[183].
XOptimal switching sequence with modulator in its formulation [184].

conventional ones (for a two-level converter). Thus, adjacent [185] which facilitates mode-transitions without an islanding
virtual voltage vectors have smaller spacing between them, detection algorithm. It also allows fault ride-through operation
giving much lower current ripples. This method also has lower with protection from excessive fault currents. An improved
computational requirements than conventional MPC. MPC-based mode-transition scheme is reported in [186] with
Fig. 11 illustrates the multiple-vector-based predictive a unified predictive voltage and current cost function. This
power control of a grid-connected converter [172]. It utilizes improves the smoothness of transition from voltage control
the computation of instantaneous power for active and reactive (when the converter is functioning as a grid-forming converter)
power slopes within each sampling cycle. These provide inputs to current control (when functioning as a grid-following con-
for the determiniation of the reference voltage vectors and their verter).
corresponding on and off time durations (duty cycles). Space
vector modulation (SVM) is finally applied for the pulse-width V. A RTIFICIAL I NTELLIGENCE -BASED P REDICTIVE
modulation of the converter. C ONTROL
Fig. 12 shows the direct predictive power control of a Data-based methods are facilitated by the plethora of data
three-level neutral-point clamped converter based on optimal that is generated by sensors, IoT, edge computing, digital
switching sequence. It comprises a cascade of outer MPC for twin and big data analytics. These data serve as inputs for
power control, and inner MPC for DC-link capacitor control. artificial-intelligence (AI) design optimization, control, and
First, a relaxed solution of voltage vectors is obtained within real-time condition monitoring of power electronics [187]. AI-
specified constraints. Next, an optimal sector search is done based MPC methods emerged from the necessity to enhance
to determined the optimal sector (comprising three voltage the performance of MPC in three main areas: easier optimal
vectors in [183]). A reduced region is delimited within the tuning of weighting factors (and prediction horizon), reduction
optimal sector, and then optimal duty cyle is calculated for of the computational burden, and parametric estimation. Two
application to the converter with optimal switching sequence. broad classifications of these techniques are shown in Fig.
13, namely, AI-type-based classification and purpose-based
C. Mode Transition for Microgrid-Based DERs classification. Considering the type of AI-method involved,
The functions of a microgrid require it to operate in both there are six methods: artificial neural network (ANN), fuzzy
standalone and grid-connected modes. Thus, it is essential logic, deep learning, reinforcement learning, particle swarm
to regulate its transition from standalone to grid-connected optimization and neuro-fuzzy logic. ANN is the most popular
operation and vice-versa. Poor control in this period can among these methods for 30 total papers reviewed (Table IV).
result in loss of synchronism, voltage/current overshoots, and
even instability. Universal controllers are utilized to achieve The purpose-based classification identifies twelve purposes
this goal to ensure seamless switching between different (or objectives) associated with AI-MPC methods. The most
operational modes. A unified control scheme is presented in recurrent in the literature include (in decreasing popularity)
 0
0 10 20 30 40 50 13
Harmonic order

6
Grid Grid
5

4
DC-AC
[%]

DC-AC
3
PCCn

abc abc abc abc

PLL
voltage vectors

2
i

SVM SVM
Actual

1
Ref. & virtual

voltage vectors Ref. voltage

Instantaneous

0
0 10 20 30 40 50 calculation power calc.
Min.
Harmonic order Model

Cost Function
sc < 20
(a) k prediction
Duty cycle
Power slopes
10 calculation calculation
PI

8
(a) (b)
Fig. 11. Predictive control of grid-connected converters with fixed switching
6
[%]

frequency: multiple-vector-based predictive power control [172].

PCCn

4
i

Grid

DC-AC
0
0 10 20 30 40 50 abc abc
Harmonic order Inner 

MPC
(b) 20 < ksc < 50
Fig. 9.6 Current harmonic limits at the point of common coupling for systems
Optim. &

rated 120 V through 69 kV (IEEE 519 − 2014 standard) [165]. ( ksc is the duty cycle
Preprocessing
Outer

5
short circuit ratio.) MPC

4 Region
Optimal

search sector search

[%]

Grid Grid
3
PCCn

Fig. 12. Direct predictive power control based on optimal switching sequence
2
i

[183].
DC-AC DC-AC
1
abc abc abc abc
PLL
voltage vectors

0 [189], MMC [190]–[192], and resonant power converter [193].

SVM SVM
0 10 20 30 40 50
Actual

These solutions involve the offline training of an ANN model

Harmonic order (and deep learning) to achieve identical transient and steady-
Ref. & virtual

voltage vectors
state performance as MPC. They have the advantage of lower
Ref. voltage
Instantaneous

memory computational burden,power

calculation and faster
calc. implementation time
Min.
Model
than the comparable MPC which they emulate. ANN-based
Cost Function
prediction offline weighting factor design involves training a surrogate
modelDutyofcycle
Power slopesor experimental data.
the system from simulation
calculation calculation
Then, a user-defined fitness function gives a multi-objective
PI
optimization of weighting factor – more accurate than heuristic
(a) tuning. Applications (b)
include VSCs [194], [195], dual active-
Fig. 10. Predictive control of grid-connected converters with fixed switching
bridge (DAB) converter for aircraft microgrid [195], and in-
frequency: Floating virtual voltage-vector-based predictive control [171]. duction motor [196]. AI-based online weighting factor tuning
involves a real-time dynamic update of optimal weighting
factors in the cost function. This results in improved reference
online weighting factor tuning, emulation of MPC algorithm, tracking accuracy under varying operational conditions for
offline weighting factor design and condition monitoring (see VSCs [197], [198], three-level NPC converter [199], and
Table IV). AI methods which emulate MPC’s optimal control PMSM [200].
find application in two-level voltage source converters [188], Several other control objectives are reported in the litera-
 14

Artificial-Intelligence-Based Predictive Control

AI-Type-based classification Purpose-based classification

Artificial neural network Fuzzy logic MPC algorithm emulation Online parameter estimation

Deep learning Particle swarm optimization Offline weighting factor design Blackbox data-driven controller

Reinforcement learning Neuro-fuzzy logic Online weighting factor tuning Load modelling

Model-parameter-free intelligent control Converter impedance estimation

Cyber-attack detection and mitigation Event-triggered MPC

Online optimization of prediction horizon Condition monitoring

Fig. 13. Broad classifications of AI-MPC-based control methods.

TABLE IV cyber-physical interaction between the flexible sources and

P OPULARITY OF AI- BASED MPC METHODS optimal controllers, and is facilitated by electricity market
stakeholders [221].
Method/Purpose Frequency
(%) Studies on MPC-based VPP energy management and price
AI-Type Artificial neural network 56.7 arbitrage indicate promising opportunities. Energy-arbitrage
Fuzzy logic 26.7 was actualized in [222], by centralized MPC, for a VPP from
Deep learning 3.3 battery storage, PC and diesel generators. Authors in [223]
Reinforcement learning 3.3
implemented optimal energy-price arbitrate with distributed
Regression-based learning 3.3
Neuro-fuzzy logic 3.3 MPC for multiple sources and thermal management. Optimal
Particle swarm optimization 3.3 day-ahead scheduling is computed with MPC for a VPP
Total* 100.0 with several grid support services [224], as shown in figure
Control purpose Online weighting factor tuning 40.0 14. Other studies consider nonlinear battery aging [225],
Emulation of MPC 16.7
frequency regulation [226], and real-time operations [227],
Offline weighting factor design 10.0
Condition monitoring 6.7 [228].
Others** 26.6
Total* 100.0
* Sum may not equal 100% due to rounding errors. ** Others B. Grid Ancillary Services
include model-parameter-free intelligent control, cyber-attack
detection and mitigation, event-triggered MPC, online param- Ancillary services are specialty services that facilitate reli-
eter estimation, blackbox data-driven controller, load mod- able power supply in the grid. They are supplied by specialty
elling, converter impedance estimation, online optimization
of prediction horizon. providers to the system operator [27], and include frequency
control, voltage and reactive power control, black start capa-
bility, oscillation damping, congestion management and loss
compensation (see Fig. 15). Among these, the most commonly
ture. These include: model-parameter-free control [201], false- required are frequency and voltage control services. Frequency
data injection cyber-attack detection and mitigation [202], control involves maintaining the frequency at regulatory levels
event-triggered MPC [203], online parameter estimation [204], by ensuring a balance between active power generated and
blackbox data-driven control [205], load modelling [206], consumed [27]. The deployment of positive and negative
converter terminal impedance estimation [207], and online frequency control reserves helps to achieve this goal. Voltage
optimization of prediction horizon [208]. Further details on control service regulates participating devices which generate
the control objectives are provided in Table V. or absorb reactive power as a means to control voltage levels
[28]–[30]. As the smart grid accommodates an increasing num-
VI. V IRTUAL P OWER P LANTS AND G RID A NCILLARY ber of converter-interfaced DERs, MPC becomes beneficial to
S ERVICES implement optimal frequency [31], [32] and voltage control
In this section, the predictive control of grid ancillary [33], [34]. The predictive control of several heterogeneous
services and virtual power plants will be discussed. thermostatically-controlled loads to provide ancillary service
was validated in [35]. It was shown in [8] that by engaging
ancillary services within an MPC framework, the microgrid
A. Virtual Power Plants operating expenses can be reduced by almost 25%.
A virtual power plant (VPP) is an aggregator of spatially As modern power grids become more renewable and
distributed energy resources for present or future grid (or environment-friendly, they also require inertia support. In
microgrid) management [219]. The energy sources in a VPP particular, replacing fossil-powered synchronous generators
are physically located at/within [220] DERs, microgrids, build- with converter-interfaced renewable sources comes at a price
ings, and PEVs. The virtual coordination is carried out through of reduced system inertia. The virtual synchronous generator
 15

TABLE V
OVERVIEW OF AI- BASED PREDICTIVE CONTROL APPLICATIONS TO POWER CONVERTERS .

Purpose Method Feature Advantage Application

Emulation of MPC XANN XANN-based emulation of MPC algo- XLower computational burden. X2L VSC [188], [189]
XDeep learn- rithm. XMMC [190]–[192]
ing XIdentical transient and steady-state Xresonant power con-
performance as MPC. verter [193]
XOffline-trained emulator.
Offline weighting factor XANN XTrains offline surrogate model of the XMore accurate weighting factor is de-
XVSC [194], [195]
design system from model data. rived than heuristic tuning. XDAB converter air-
XUser-defined fitness function gives XImproved dynamic performance of craft microgrid [195]
multi-objective optimization of weight- system. Xinduction motor
ing factor. [196]
Online weighting factor XReinforcement XAutomated online training. XImproves reference tracking accuracy XVSC [197], [198]
tuning learning XOnline dynamic update of optimal under varying operational conditions. X3L NPC converter
XANN weighting factor. XEasier weighting factor tuning. [199], [209]
XPSO Xdirect matrix con-
XFuzzy logic verter [210]
Xmachine drives
[200], [211]–[216]
Model-parameter-free XANN XCascaded predictor-based neural net- XModel-free weighting factor-free con- XMMC [201]
control work for system identification. trol.
XImproved robustness to parametric un-
certainties.
False data injection XANN XANN controller for cyber-attack de- XImpact of cyber-attack on the micro- XDC Microgrids [202]
cyber-attack detection tection and mitigation. grid is mitigated.
and mitigation
Event triggered MPC XANN XNN controller trained to emulate XRobustness to uncertainties and en- XMMC [203]
event-triggered MPC. ergy loss minimization.
XLow switching frequency control for
MMC.
Online parameter esti- XNeuro-fuzzy XNeuro-fuzzy logic-based online model XImproved reference tracking accuracy. X2L VSC [204]
mation logic parameter estimation. XImproved robustness to parametric
variations.
Blackbox data-driven XRegression- XCalculation of conditional entropy for XImproved classification accuracy. XVSC [205]
control based learning input-output mapping. XIll-impact of corrupted data in training
XErroneous data detection and process- is eliminated.
ing.
Load modelling XANN XData-based surrogate model of load. XImproved prediction accuracy. XMatrix converter
[206]
Converter terminal XANN XConverter parameter (impedance) XImproved reference tracking accuracy. XVSC [207]
impedance estimation identication through ANN.
XLearned impedance factor added to
MPC cost function.
Online optimization of XANN XOnline calculation of optimal predic- XImproved reference tracking accuracy. XBoost converter
prediction horizon tion horizon. XImproved robustness to parametric [208]
XPrediction horizon adapts to the oper- variations.
ational states of the converter.
Condition monitoring XANN XFault detection by wavelet trans- XImproved accuracy of fault detection. XGrid-connected PV
form and neural network-based island- XAccurate fault identification. system [217]
ing classifier. X3L NPC converter
XLVRT during voltage sags. [218]
XOpen-circuit fault diagnosis.

(VSG) has become beneficial to support grid frequency re- is essential to maximize financial returns from such capital
sponse and control. MPC-based VSG control is reported in intensive projects.
[229], and it is adaptive to electrical load conditions. Fault Congestion management in transmission [235] and distri-
ride-through and over-current protection for VSG was studied bution [236] networks can be achieved by optimal control
in [230]. The application of VSG with MPC-based converter of ESS and partial curtailment of renewable sources. The
in [231], [232] gives better rate of change of frequency (within study in [237] proposed a distributed model predictive control
regulatory levels) and system stability than conventional droop solution for economic dispatch of DER. The method utilizes
control. In addition, ESS-supported fast frequency response both forecast data and stochastic variables in the prediction
for networked microgrids is also achievable through MPC model. A similar problem was solved in [238] for DER in a
and multi-agent control theory [233]. Practical projects have microgrid.
been implemented that deployed grid-scale battery storage for Voltage support in a microgrid was investigated in [239],
frequency regulation ancillary services; an example was re- [240]. Similarly, bus voltages in a DC-microgrid can be sup-
ported by East Penn, USA [234]. Energy storage optimization ported with controlled rate of change of voltage [241]. Voltage
 16

Day-ahead scheduler Legend

Arbitrage profile Day-ahead
References
Frequency regulation market prices Measurements

Model Predictive Controller

Thermal

Wind Solar PV Building storage

Battery
CHP
PEV

Fig. 14. Model predictive control of virtual power plants [224].

Ancillary Services
Tert. response
Sec. control

> 15 mins (distributed)

Sec. response
Sec. control

30 s to 15 mins Voltage &

(decentralized)
Frequency
Black start
Oscillation
Congestion
Loss

reactive power
Prim. response
control capability damping management compensation   Sec. control   

control
< 30 s (centralized)
Inertial response
Prim. control
< 10 s

Fig. 15. Ancillary services for the smart grid [243], [244].

control can be achieved in centralized [16], decentralized distribution, multi-fuel inputs, and multi-service applications
[242], and distributed [232] topologies. Centralized topology [248]. Whereas energy systems were historically controlled as
requires more sophisticated communication resources than independent agents, more recent requirements of economical
distributed methods. Distributed control is more effective for minimization of environmental impacts demand their synergis-
wide area electrical networks [18], [232]. tic manipulation. This implies a more intricate simultaneous
optimization of multiple energy vectors (e.g., electricity and
VII. O PEN I SSUES AND F UTURE T RENDS heat) in an interactive manner. A few ground-breaking studies
This section will review open issues and future perspectives that adopt this philosophy show that the MPC-based multi-
on the development of predictive control of DERs. energy approach can improve the economic operation of DERs
[224], [249].
A. Open Research Issues Interconnected multi-microgrids have the potential to en-
Despite the highly promising characteristics of predictive hance the reliability of power supply through the sharing
control for distributed generation, it has several limitations, of energy resources among nearby microgrids. Therefore,
and two key issues will be discussed here. First, MPC’s when a microgrid has a critical shortage in power supply,
performance is strongly determined by the accuracy of sys- a better alternative to load-shedding will be power-sharing
tem’s mathematical model. Therefore, external disturbances by a neighbor-microgrid (which has excess power). MPC
e.g., stochastic parameteric variations can deteriorate the con- can help in this regard to optimize power sharing subject
trol performance. Recent research solutions to this challenge to multiple objectives, and operational constraints. This will
include: adaptive MPC with revised prediction model [69], reduce energy losses, and extend the lifespan of energy storage
[245], and model-free predictive control [246]. Second, MPC’s systems. In addition, MPC-based methods can also provide
ability to control multiple objectives is enhanced through other ancillary services like frequency and voltage control to
weighting factors. Nonetheless, calculation of optimal weight- microgrids within a common cluster.
ing factors can be laborious. Some solutions to this include
Data-science and artificial intelligence techniques are ex-
optimization techniques for offline or online weighting factor
pected to improve the cost-effective operation of large numbers
computation, and MPC without weighting factors [247].
of DERs in the smart grid. Thus, data-based methods are
expected to further simplify, and improve the industrial appli-
B. Future Trends cability of predictive control. For instance, low-computational-
The proliferation of DERs creates a need for innovative resource intelligent algorithms that emulate advanced MPC
planning, and optimal real-time operation of multi-energy with long prediction horizons, and model-free data-driven
systems. Multi-energy systems are characterized by spatial methods.
 17

VIII. C ONCLUSION [16] K. T. Tan, X. Y. Peng, P. L. So, Y. C. Chu, and M. Z. Chen, “Centralized
control for parallel operation of distributed generation inverters in
The stringent grid-code requirements for grid-connected microgrids,” IEEE Transactions on Smart Grid, vol. 3, no. 4, pp. 1977–
distributed energy resources (DERs) necessitate high perfor- 1987, 2012.
[17] J. Liu, Y. Miura, and T. Ise, “Cost-function-based microgrid decen-
mance, multi-objective control methods for power electronic tralized control of unbalance and harmonics for simultaneous bus
converters. In this article, the applications of MPC to the smart voltage compensation and current sharing,” IEEE Transactions on
grid were introduced. Furthermore, a comprehensive review Power Electronics, vol. 34, pp. 7397–7410, 2019.
[18] K. Liu, T. Liu, Z. Tang, and D. J. Hill, “Distributed MPC-Based
was done on power converters for wind energy conversion Frequency Control in Networked Microgrids With Voltage Constraints,”
systems (WECS), solar photovoltaic, fuel cell, and energy stor- IEEE Transactions on Smart Grid, vol. 10, no. 6, pp. 6343–6354, mar
age systems. Complementing MPC with artificial intelligence 2019.
[19] R. Heydari, Y. Khayat, A. Amiri, T. Dragicevic, Q. Shafiei, P. Popovsky,
offers benefits including: lower computational burden, easier and F. Blaabjerg, “Robust high-rate secondary control of microgrids
and more accurate weighting factor tuning, improved reference with mitigation of communication impairments,” IEEE Transactions
tracking accuracy, and improved robustness to parametric un- on Power Electronics, 2020.
[20] F. Garcia-Torres, L. Valverde, and C. Bordons, “Optimal Load Sharing
certainties. The future trends of MPC show its good potential of Hydrogen-Based Microgrids with Hybrid Storage Using Model-
to support emerging technologies, viz., multi-energy systems, Predictive Control,” IEEE Transactions on Industrial Electronics,
multi-microgrids, and virtual power plants. vol. 63, no. 8, pp. 4919–4928, 2016.
[21] F. Garcia-Torres, C. Bordons, and M. A. Ridao, “Optimal economic
schedule for a network of microgrids with hybrid energy storage
R EFERENCES system using distributed model predictive control,” IEEE Transactions
on Industrial Electronics, vol. 66, no. 3, pp. 1919–1929, 2019.
[1] X. Yu and Y. Xue, “Smart grids: A cyber-physical systems perspective,” [22] T. Morstyn, B. Hredzak, and V. G. Agelidis, “Network Topology
Proceedings of the IEEE, vol. 104, pp. 1058–1070, 2016. Independent Multi-Agent Dynamic Optimal Power Flow for Microgrids
[2] G. W. Arnold, “Challenges and opportunities in smart grid: A position with Distributed Energy Storage Systems,” IEEE Transactions on Smart
article,” Proceedings of the IEEE, vol. 99, pp. 922–927, 2011. Grid, vol. 9, no. 4, pp. 3419–3429, 2018.
[3] Z. Zhang, O. Babayomi, T. Dragicevic, R. Heydari, C. Garcia, J. Ro- [23] Z. Guo, H. Jiang, Y. Zheng, and S. Li, “Distributed model predictive
driguez, and R. Kennel, “Advances and opportunities in the model control for efficient operation of islanded microgrid,” in 2017 Chinese
predictive control of microgrids: Part i–primary layer,” International Automation Congress (CAC), vol. 2017-Janua, IEEE. Institute of
Journal of Electrical Power Energy Systems, vol. 134, p. 107411, Electrical and Electronics Engineers Inc., dec 2017, pp. 6253–6258.
2022. [24] R. Deng, Z. Yang, M.-Y. Chow, and J. Chen, “A survey on demand
[4] O. Babayomi, Z. Zhang, T. Dragicevic, R. Heydari, Y. Li, C. Garcia, response in smart grids: Mathematical models and approaches,” IEEE
J. Rodriguez, and R. Kennel, “Advances and opportunities in the model Transactions on Industrial Informatics, vol. 11, pp. 570–582, 2015.
predictive control of microgrids: Part ii–secondary and tertiary layers,” [25] N. Karthikeyan, J. R. Pillai, B. Bak-Jensen, and J. W. Simpson-Porco,
International Journal of Electrical Power Energy Systems, vol. 134, “Predictive control of flexible resources for demand response in active
p. 107339, 2022. distribution networks,” IEEE Transactions on Power Systems, vol. 34,
[5] J. Drgoňa, J. Arroyo, I. Cupeiro Figueroa, D. Blum, K. Arendt, D. Kim, pp. 2957–2969, 2019.
E. P. Ollé, J. Oravec, M. Wetter, D. L. Vrabie, and L. Helsen, “All you [26] M. D. Knudsen and S. Petersen, “Demand response potential of model
need to know about model predictive control for buildings,” Annual predictive control of space heating based on price and carbon dioxide
Reviews in Control, vol. 50, pp. 190–232, 2020. intensity signals,” Energy and Buildings, vol. 125, pp. 196–204, 2016.
[6] F. Lauro, F. Moretti, A. Capozzoli, and S. Panzieri, “Model predictive [27] Y. G. Rebours, D. S. Kirschen, M. Trotignon, and S. Rossignol, “A
control for building active demand response systems,” Energy Procedia, survey of frequency and voltage control ancillary services—part i:
vol. 83, pp. 494–503, 2015, sustainability in Energy and Buildings: Technical features,” IEEE Transactions on Power Systems, vol. 22,
Proceedings of the 7th International Conference SEB-15. no. 1, pp. 350–357, 2007.
[7] G. Serale, M. Fiorentini, A. Capozzoli, D. Bernardini, and A. Bem- [28] A. R. Malekpour, A. M. Annaswamy, and J. Shah, “Hierarchical hybrid
porad, “Model predictive control (mpc) for enhancing building and architecture for volt/var control of power distribution grids,” IEEE
hvac system energy efficiency: Problem formulation, applications and Transactions on Power Systems, vol. 35, no. 2, pp. 854–863, 2020.
opportunities,” Energies, vol. 11, no. 3, 2018. [29] T. M. S. Ibrahim, T. T. De Rubira, A. Del Rosso, M. Patel, S. Guggilam,
[8] J. R. Nelson and N. G. Johnson, “Model predictive control of mi- and A. Mohamed, “Alternating optimization approach for voltage-
crogrids for real-time ancillary service market participation,” Applied secure multi-period optimal reactive power dispatch,” IEEE Transac-
Energy, vol. 269, p. 114963, 2020. tions on Power Systems, pp. 1–1, 2021.
[9] B. Varga, S. Meier, S. Schwab, and S. Hohmann, “Model predictive [30] Z. Tang, D. J. Hill, and T. Liu, “Distributed coordinated reactive
control and trajectory optimization of large vehicle-manipulators,” in power control for voltage regulation in distribution networks,” IEEE
2019 IEEE International Conference on Mechatronics (ICM), vol. 1. Transactions on Smart Grid, vol. 12, no. 1, pp. 312–323, 2021.
IEEE, 2019, pp. 60–66. [31] Y. Jia, Z. Y. Dong, C. Sun, and K. Meng, “Cooperation-based dis-
[10] T. Geyer, Model predictive control of high power converters and tributed economic mpc for economic load dispatch and load frequency
industrial drives. John Wiley & Sons, 2016. control of interconnected power systems,” IEEE Transactions on Power
[11] R. Wang, G. Xiao, and P. Wang, “Hybrid centralized-decentralized Systems, vol. 34, no. 5, pp. 3964–3966, 2019.
(hcd) charging control of electric vehicles,” IEEE Transactions on [32] Z. Yi, Y. Xu, W. Gu, and Z. Fei, “Distributed model predictive control
Vehicular Technology, vol. 66, no. 8, pp. 6728–6741, 2017. based secondary frequency regulation for a microgrid with massive dis-
[12] Y. Shi, H. D. Tuan, A. V. Savkin, T. Q. Duong, and H. V. Poor, tributed resources,” IEEE Transactions on Sustainable Energy, vol. 12,
“Model predictive control for smart grids with multiple electric-vehicle no. 2, pp. 1078–1089, 2021.
charging stations,” IEEE Transactions on Smart Grid, vol. 10, no. 2, [33] G. Lou, W. Gu, W. Sheng, X. Song, and F. Gao, “Distributed model
pp. 2127–2136, 2019. predictive secondary voltage control of islanded microgrids with feed-
[13] Y. Shan, J. Hu, M. Liu, J. Zhu, and J. M. Guerrero, “Model Predictive back linearization,” IEEE Access, vol. 6, pp. 50 169–50 178, sep 2018.
Voltage and Power Control of Islanded PV-Battery Microgrids with [34] J. S. Gómez, D. Sáez, J. W. Simpson-Porco, and R. Cárdenas, “Dis-
Washout-Filter-Based Power Sharing Strategy,” IEEE Transactions on tributed predictive control for frequency and voltage regulation in
Power Electronics, vol. 35, no. 2, pp. 1227–1238, 2020. microgrids,” IEEE Transactions on Smart Grid, vol. 11, no. 2, pp.
[14] Z. Cheng, J. Duan, and M.-Y. Chow, “To centralize or to distribute: 1319–1329, 2019.
That is the question: A comparison of advanced microgrid management [35] M. Liu and Y. Shi, “Model predictive control of aggregated hetero-
systems,” IEEE Industrial Electronics Magazine, vol. 12, no. 1, pp. 6– geneous second-order thermostatically controlled loads for ancillary
24, 2018. services,” IEEE Transactions on Power Systems, vol. 31, no. 3, pp.
[15] T. L. Vandoorn, J. D. De Kooning, B. Meersman, and L. Vandevelde, 1963–1971, 2016.
“Review of primary control strategies for islanded microgrids with [36] J. Rodriguez, C. Garcia, A. Mora, F. Flores-Bahamonde, P. Acuna,
power-electronic interfaces,” pp. 613–628, 2013. M. Novak, Y. Zhang, L. Tarisciotti, A. Davari, Z. Zhang, F. Wang,
 18

M. Norambuena, T. Dragicevic, F. Blaabjerg, T. Geyer, R. Kennel, [55] W. Liu and Y. Liu, “Hierarchical model predictive control of wind farm
D. Arab Khaburi, M. Abdelrahem, Z. Zhang, N. Mijatovic, and R. P. with energy storage system for frequency regulation during black-start,”
Aguilera, “Latest advances of model predictive control in electrical International Journal of Electrical Power Energy Systems, vol. 119,
drives. part i: Basic concepts and advanced strategies,” IEEE Transac- p. 105893, 2020.
tions on Power Electronics, pp. 1–1, 2021. [56] B. E. Sedhom, M. M. El-Saadawi, A. Y. Hatata, and A. S. Alsayyari,
[37] R. Baidya, R. P. Aguilera, P. Acuña, S. Vazquez, and H. d. T. Mouton, “Hierarchical control technique-based harmony search optimization
“Multistep model predictive control for cascaded h-bridge inverters: algorithm versus model predictive control for autonomous smart mi-
Formulation and analysis,” IEEE Transactions on Power Electronics, crogrids,” International Journal of Electrical Power Energy Systems,
vol. 33, no. 1, pp. 876–886, 2018. vol. 115, p. 105511, 2020.
[38] P. Acuna, C. A. Rojas, R. Baidya, R. P. Aguilera, and J. E. Fletcher, [57] O. Babayomi, Y. Li, and Z. Zhang, “Distributed consensus-based
“On the impact of transients on multistep model predictive control reactive power sharing in microgrids: A predictive virtual capacitance
for medium-voltage drives,” IEEE Transactions on Power Electronics, control technique,” International Journal of Electrical Power Energy
vol. 34, no. 9, pp. 8342–8355, 2019. Systems, vol. 141, p. 108139, 2022.
[39] Y. Shi, H. D. Tuan, A. V. Savkin, T. Q. Duong, and H. V. Poor, [58] Y. Xie, L. Liu, Q. Wu, and Q. Zhou, “Robust model predictive
“Model predictive control for smart grids with multiple electric-vehicle control based voltage regulation method for a distribution system with
charging stations,” IEEE Transactions on Smart Grid, vol. 10, no. 2, renewable energy sources and energy storage systems,” International
pp. 2127–2136, 2019. Journal of Electrical Power Energy Systems, vol. 118, p. 105749,
[40] S. Bansal, M. N. Zeilinger, and C. J. Tomlin, “Plug-and-play model 2020.
predictive control for electric vehicle charging and voltage control in [59] P. Karamanakos, E. Liegmann, T. Geyer, and R. Kennel, “Model
smart grids,” in 53rd IEEE Conference on Decision and Control, 2014, Predictive Control of Power Electronic Systems: Methods, Results, and
pp. 5894–5900. Challenges,” IEEE Open Journal of Industry Applications, vol. 1, pp.
[41] A. Di Giorgio, F. Liberati, and S. Canale, “Electric vehicles charging 95–114, 2020.
control in a smart grid: A model predictive control approach,” Control [60] K.-W. Hu and C.-M. Liaw, “Development of a wind interior permanent-
Engineering Practice, vol. 22, no. Complete, pp. 147–162, 2014. magnet synchronous generator-based microgrid and its operation con-
[42] N. I. Nimalsiri, C. P. Mediwaththe, E. L. Ratnam, M. Shaw, D. B. trol,” IEEE Transactions on Power Electronics, vol. 30, no. 9, pp. 4973–
Smith, and S. K. Halgamuge, “A survey of algorithms for distributed 4985, 2015.
charging control of electric vehicles in smart grid,” IEEE Transactions [61] S. Li and J. Li, “Output predictor-based active disturbance rejection
on Intelligent Transportation Systems, vol. 21, no. 11, pp. 4497–4515, control for a wind energy conversion system with pmsg,” IEEE Access,
2020. vol. 5, pp. 5205–5214, 2017.
[43] V. Saxena, N. Kumar, B. Singh, and B. K. Panigrahi, “An mpc based [62] M. Abdelrahem, C. M. Hackl, and R. Kennel, “Finite position set-phase
algorithm for a multipurpose grid integrated solar pv system with locked loop for sensorless control of direct-driven permanent-magnet
enhanced power quality and pcc voltage assist,” IEEE Transactions synchronous generators,” IEEE Transactions on Power Electronics,
on Energy Conversion, vol. 36, no. 2, pp. 1469–1478, 2021. vol. 33, no. 4, pp. 3097–3105, 2018.
[44] Z. Cui, Z. Zhang, T. Dragicevi?, and J. Rodrı́guez, “Dynamic Sequen- [63] J.-S. Lee and K.-B. Lee, “Predictive control of vienna rectifiers for
tial Model Predictive Control of Three-Level NPC Back-to-Back Power pmsg systems,” IEEE Transactions on Industrial Electronics, vol. 64,
Converter PMSG Wind Turbine Systems,” in IECON 2020 The 46th no. 4, pp. 2580–2591, 2017.
Annual Conference of the IEEE Industrial Electronics Society, 2020, [64] A. Calle-Prado, S. Alepuz, J. Bordonau, P. Cortes, and J. Rodriguez,
pp. 3206–3211, iSSN: 2577-1647. “Predictive control of a back-to-back npc converter-based wind power
[45] B. Liu, G. Li, D. He, and Y. Chen, “Dc and ac power quality system,” IEEE Transactions on Industrial Electronics, vol. 63, no. 7,
control for single-phase grid-tied pemfc systems with low dc-link pp. 4615–4627, 2016.
capacitance by solution-space-reduced mpc,” IEEE Transactions on [65] M. Abdelrahem, C. M. Hackl, Z. Zhang, and R. Kennel, “Robust
Industrial Electronics, pp. 1–1, 2021. predictive control for direct-driven surface-mounted permanent-magnet
[46] A. Lashab, D. Sera, J. M. Guerrero, L. Mathe, and A. Bouzid, “Discrete synchronous generators without mechanical sensors,” IEEE Transac-
model-predictive-control-based maximum power point tracking for pv tions on Energy Conversion, vol. 33, no. 1, pp. 179–189, 2018.
systems: Overview and evaluation,” IEEE Transactions on Power [66] R. Errouissi, A. Al-Durra, S. M. Muyeen, S. Leng, and F. Blaabjerg,
Electronics, vol. 33, no. 8, pp. 7273–7287, 2018. “Offset-free direct power control of dfig under continuous-time model
[47] N. T. Mbungu, R. M. Naidoo, R. C. Bansal, and V. Vahidinasab, predictive control,” IEEE Transactions on Power Electronics, vol. 32,
“Overview of the optimal smart energy coordination for microgrid no. 3, pp. 2265–2277, 2017.
applications,” IEEE Access, vol. 7, pp. 163 063–163 084, 2019. [67] A. J. Sguarezi Filho, A. L. de Oliveira, L. L. Rodrigues, E. C. M.
[48] N. T. Mbungu, R. Naidoo, R. C. Bansal, and M. W. Siti, “Model Costa, and R. V. Jacomini, “A robust finite control set applied to the
predictive control: A survey of dynamic energy management.” in dfig power control,” IEEE Journal of Emerging and Selected Topics in
ICINCO, 2021, pp. 123–129. Power Electronics, vol. 6, no. 4, pp. 1692–1698, 2018.
[49] D. Razmi, O. Babayomi, A. Davari, T. Rahimi, Y. Miao, and Z. Zhang, [68] Z. Zhang, H. Fang, F. Gao, J. Rodrı́guez, and R. Kennel, “Multiple-
“Review of model predictive control of distributed energy resources in Vector Model Predictive Power Control for Grid-Tied Wind Turbine
microgrids,” Symmetry, vol. 14, no. 8, p. 1735, 2022. System With Enhanced Steady-State Control Performance,” IEEE
[50] F. Ni, Z. Zheng, Q. Xie, X. Xiao, Y. Zong, and C. Huang, “Enhancing Transactions on Industrial Electronics, vol. 64, no. 8, pp. 6287–6298,
resilience of dc microgrids with model predictive control based hybrid 2017.
energy storage system,” International Journal of Electrical Power [69] Z. Zhang, Z. Li, M. P. Kazmierkowski, J. Rodriguez, and R. Kennel,
Energy Systems, vol. 128, p. 106738, 2021. “Robust Predictive Control of Three-Level NPC Back-to-Back Power
[51] M. Cao, Q. Xu, X. Qin, and J. Cai, “Battery energy storage sizing based Converter PMSG Wind Turbine Systems with Revised Predictions,”
on a model predictive control strategy with operational constraints to IEEE Transactions on Power Electronics, vol. 33, no. 11, pp. 9588–
smooth the wind power,” International Journal of Electrical Power 9598, nov 2018.
Energy Systems, vol. 115, p. 105471, 2020. [70] M. E. Zarei, C. Veganzones Nicolás, and J. Rodrı́guez Arribas,
[52] P. C. Blaud, P. Haurant, F. Claveau, B. Lacarrière, P. Chevrel, and “Improved predictive direct power control of doubly fed induction
A. Mouraud, “Modelling and control of multi-energy systems through generator during unbalanced grid voltage based on four vectors,” IEEE
multi-prosumer node and economic model predictive control,” Inter- Journal of Emerging and Selected Topics in Power Electronics, vol. 5,
national Journal of Electrical Power Energy Systems, vol. 118, p. no. 2, pp. 695–707, 2017.
105778, 2020. [71] M. E. Zarei, C. V. Nicolás, J. R. Arribas, and D. Ramı́rez, “Four-switch
[53] L. Bartolucci, S. Cordiner, V. Mulone, and J. L. Rossi, “Hybrid renew- three-phase operation of grid-side converter of doubly fed induction
able energy systems for household ancillary services,” International generator with three vectors predictive direct power control strategy,”
Journal of Electrical Power Energy Systems, vol. 107, pp. 282–297, IEEE Transactions on Industrial Electronics, vol. 66, no. 10, pp. 7741–
2019. 7752, 2019.
[54] Y. Zhang, P. Kou, L. Yu, and D. Liang, “Coordinated voltage and [72] Y. Zhang, J. Jiao, D. Xu, D. Jiang, Z. Wang, and C. Tong, “Model pre-
frequency control for hvdc sending end under pole-block fault: Using dictive direct power control of doubly fed induction generators under
model predictive control,” International Journal of Electrical Power balanced and unbalanced network conditions,” IEEE Transactions on
Energy Systems, vol. 136, p. 107655, 2022. Industry Applications, vol. 56, no. 1, pp. 771–786, 2020.
 19

[73] Z. Zhang, C. M. Hackl, and R. Kennel, “Computationally efficient [92] V. Jayan and A. M. Ghias, “A weighting factor free model predictive
dmpc for three-level npc back-to-back converters in wind turbine control for a flying capacitor converter in a dc microgrid,” IEEE
systems with pmsg,” IEEE Transactions on Power Electronics, vol. 32, Transactions on Energy Conversion, pp. 1–1, 2021.
pp. 8018–8034, 2017. [93] V. Jayan and A. M. Y. M. Ghias, “A single-objective modulated model
[74] M. Abdelrahem, C. M. Hackl, R. Kennel, and J. Rodrı́guez, “Compu- predictive control for a multilevel flying-capacitor converter in a dc
tationally efficient finite-position-set-phase-locked loop for sensorless microgrid,” IEEE Transactions on Power Electronics, vol. 37, no. 2,
control of pmsgs in wind turbine applications,” IEEE Transactions on pp. 1560–1569, 2022.
Power Electronics, vol. 36, no. 3, pp. 3007–3016, 2021. [94] J. Chen, Y. Chen, L. Tong, L. Peng, and Y. Kang, “A backpropaga-
[75] M. Abdelrahem, C. M. Hackl, R. Kennel, and J. Rodriguez, “Efficient tion neural network-based explicit model predictive control for dc–dc
direct-model predictive control with discrete-time integral action for converters with high switching frequency,” IEEE Journal of Emerging
pmsgs,” IEEE Transactions on Energy Conversion, vol. 34, no. 2, pp. and Selected Topics in Power Electronics, vol. 8, no. 3, pp. 2124–2142,
1063–1072, 2019. 2020.
[76] J.-S. Lee, K.-B. Lee, and F. Blaabjerg, “Predictive control with discrete [95] M. A. A. Mohamed, Q. Guan, and M. Rashed, “Control of dc-dc
space-vector modulation of vienna rectifier for driving pmsg of wind converter for interfacing supercapcitors energy storage to dc micro
turbine systems,” IEEE Transactions on Power Electronics, vol. 34, grids,” in 2018 IEEE International Conference on Electrical Systems
no. 12, pp. 12 368–12 383, 2019. for Aircraft, Railway, Ship Propulsion and Road Vehicles International
[77] P. Kou, D. Liang, J. Li, L. Gao, and Q. Ze, “Finite-control-set Transportation Electrification Conference (ESARS-ITEC), 2018, pp. 1–
model predictive control for dfig wind turbines,” IEEE Transactions on 8.
Automation Science and Engineering, vol. 15, no. 3, pp. 1004–1013, [96] C. Gong, J. Lin, D. Huang, and Z. Wang, “Adrc amp; mpc based control
2018. strategy of bidirectional buck-boost converter in distributed energy
[78] G. F. Gontijo, T. C. Tricarico, B. W. França, L. F. da Silva, E. L. storage systems,” in 2021 6th International Conference on Power and
van Emmerik, and M. Aredes, “Robust model predictive rotor current Renewable Energy (ICPRE), 2021, pp. 73–79.
control of a dfig connected to a distorted and unbalanced grid driven by [97] M. Jeong and J. Biela, “Dynamic operation of buck-boost dc-dc
a direct matrix converter,” IEEE Transactions on Sustainable Energy, converters over wide operating ranges with switching based model
vol. 10, no. 3, pp. 1380–1392, 2019. predictive control (mpc),” in 2021 23rd European Conference on Power
[79] A. Lashab, D. Sera, and J. M. Guerrero, “A dual-discrete model Electronics and Applications (EPE’21 ECCE Europe), 2021, pp. P.1–
predictive control-based mppt for pv systems,” IEEE Transactions on P.10.
Power Electronics, vol. 34, no. 10, pp. 9686–9697, 2019. [98] S. Dutta, S. Bhattacharya, and M. Chandorkar, “A novel predictive
[80] M. B. Shadmand, X. Li, R. S. Balog, and H. A. Rub, “Model phase shift controller for bidirectional isolated dc to dc converter for
predictive control of grid-tied photovoltaic systems: Maximum power high power applications,” in 2012 IEEE Energy Conversion Congress
point tracking and decoupled power control,” in 2015 First Workshop and Exposition (ECCE), 2012, pp. 418–423.
on Smart Grid and Renewable Energy (SGRE), 2015, pp. 1–6. [99] Y. Li, Z. Zhang, and R. Kennel, “A full state-variable predictive control
[81] M. B. Shadmand, M. Mosa, R. S. Balog, and H. A. Rub, “An improved of bi-directional boost converters with guaranteed stability,” in 2020
mppt technique for high gain dc-dc converter using model predictive 22nd European Conference on Power Electronics and Applications
control for photovoltaic applications,” in 2014 IEEE Applied Power (EPE’20 ECCE Europe), 2020, pp. P.1–P.7.
Electronics Conference and Exposition - APEC 2014, 2014, pp. 2993–
[100] A. T. Elsayed, T. A. Youssef, and O. A. Mohammed, “Modeling
2999.
and control of a low-speed flywheel driving system for pulsed-load
[82] M. Metry and R. S. Balog, “An adaptive model predictive controller for
mitigation in dc distribution networks,” IEEE Transactions on Industry
current sensorless mppt in pv systems,” IEEE Open Journal of Power
Applications, vol. 52, no. 4, pp. 3378–3387, 2016.
Electronics, vol. 1, pp. 445–455, 2020.
[101] Q. Xu, N. Vafamand, L. Chen, T. Dragičević, L. Xie, and F. Blaabjerg,
[83] M. Metry, M. B. Shadmand, R. S. Balog, and H. Abu-Rub, “Mppt of
“Review on advanced control technologies for bidirectional dc/dc
photovoltaic systems using sensorless current-based model predictive
converters in dc microgrids,” IEEE Journal of Emerging and Selected
control,” IEEE Transactions on Industry Applications, vol. 53, no. 2,
Topics in Power Electronics, vol. 9, no. 2, pp. 1205–1221, 2021.
pp. 1157–1167, 2017.
[84] M. Metry, M. B. Shadmand, R. S. Balog, and H. Abu Rub, “High [102] S. Yousefizadeh, J. D. Bendtsen, N. Vafamand, M. H. Khooban,
efficiency mppt by model predictive control considering load distur- T. Dragicevic, and F. Blaabjerg, “EKF-Based Predictive Stabilization of
bances for photovoltaic applications under dynamic weather condition,” Shipboard DC Microgrids with Uncertain Time-Varying Load,” IEEE
in IECON 2015 - 41st Annual Conference of the IEEE Industrial Journal of Emerging and Selected Topics in Power Electronics, vol. 7,
Electronics Society, 2015, pp. 004 092–004 095. no. 2, pp. 901–909, jun 2019.
[85] O. Abdel-Rahim and H. Funato, “Model predictive control based [103] N. Vafamand, M. H. Khooban, T. Dragičević, and F. Blaabjerg,
maximum power point tracking technique applied to ultra step-up boost “Networked Fuzzy Predictive Control of Power Buffers for Dynamic
converter for pv applications,” in 2014 IEEE Innovative Smart Grid Stabilization of DC Microgrids,” IEEE Transactions on Industrial
Technologies - Asia (ISGT ASIA), 2014, pp. 138–142. Electronics, vol. 66, no. 2, pp. 1356–1362, feb 2019.
[86] R. Errouissi, S. M. Muyeen, A. Al-Durra, and S. Leng, “Experimental [104] M. Vrlić, D. Ritzberger, and S. Jakubek, “Efficient and life preserving
validation of a robust continuous nonlinear model predictive control power tracking control of a proton exchange membrane fuel cell using
based grid-interlinked photovoltaic inverter,” IEEE Transactions on model predictive control,” in 2020 SICE International Symposium on
Industrial Electronics, vol. 63, no. 7, pp. 4495–4505, 2016. Control Systems (SICE ISCS), 2020, pp. 77–84.
[87] R. Errouissi, A. Al-Durra, and S. M. Muyeen, “A robust continuous- [105] J. Luna, E. Usai, A. Husar, and M. Serra, “Enhancing the efficiency
time mpc of a dc–dc boost converter interfaced with a grid-connected and lifetime of a proton exchange membrane fuel cell using nonlinear
photovoltaic system,” IEEE Journal of Photovoltaics, vol. 6, no. 6, pp. model-predictive control with nonlinear observation,” IEEE Transac-
1619–1629, 2016. tions on Industrial Electronics, vol. 64, no. 8, pp. 6649–6659, 2017.
[88] S. Sajadian and R. Ahmadi, “Distributed maximum power point track- [106] Y. Shan, J. Hu, K. W. Chan, Q. Fu, and J. M. Guerrero, “Model
ing using model predictive control for photovoltaic energy harvesting Predictive Control of Bidirectional DC-DC Converters and AC/DC
architectures based on cascaded power optimizers,” IEEE Journal of Interlinking Converters-A New Control Method for PV-Wind-Battery
Photovoltaics, vol. 7, no. 3, pp. 849–857, 2017. Microgrids,” IEEE Transactions on Sustainable Energy, vol. 10, no. 4,
[89] ——, “Model predictive-based maximum power point tracking for pp. 1823–1833, oct 2019.
grid-tied photovoltaic applications using a z-source inverter,” IEEE [107] M. Trifkovic, M. Sheikhzadeh, K. Nigim, and P. Daoutidis, “Modeling
Transactions on Power Electronics, vol. 31, no. 11, pp. 7611–7620, and control of a renewable hybrid energy system with hydrogen
2016. storage,” IEEE Transactions on Control Systems Technology, vol. 22,
[90] A. A. Abushaiba, S. M. M. Eshtaiwi, and R. Ahmadi, “A new no. 1, pp. 169–179, 2014.
model predictive based maximum power point tracking method for [108] M. R. Banaei and R. Alizadeh, “Simulation-based modeling and
photovoltaic applications,” in 2016 IEEE International Conference on power management of all-electric ships based on renewable energy
Electro Information Technology (EIT), 2016, pp. 0571–0575. generation using model predictive control strategy,” IEEE Intelligent
[91] V. Jayan, A. Ghias, and A. Merabet, “Fixed frequency model predictive Transportation Systems Magazine, vol. 8, no. 2, pp. 90–103, 2016.
control of three-level bi-directional flying capacitor dc-dc converter in [109] Y. Shan, J. Hu, and J. M. Guerrero, “A Model Predictive Power Control
dc microgrid,” in IECON 2019 - 45th Annual Conference of the IEEE Method for PV and Energy Storage Systems with Voltage Support
Industrial Electronics Society, vol. 1, 2019, pp. 3343–3348. Capability,” IEEE Transactions on Smart Grid, pp. 1–1, 2019.
 20

[110] T. Wang, H. Kamath, and S. Willard, “Control and optimization of [133] S. Stynski, W. Luo, A. Chub, L. G. Franquelo, M. Malinowski, and
grid-tied photovoltaic storage systems using model predictive control,” D. Vinnikov, “Utility-scale energy storage systems: Converters and
IEEE Transactions on Smart Grid, vol. 5, no. 2, pp. 1010–1017, 2014. control,” IEEE Industrial Electronics Magazine, vol. 14, pp. 32–52,
[111] H. Ritchie and M. Roser, “Energy,” Our World in Data, 2020, 2020.
https://ourworldindata.org/energy. [134] B. Hredzak, V. G. Agelidis, and M. Jang, “A model predictive control
[112] IRENA, “Renewable power generation costs in 2020,” 2021. system for a hybrid battery-ultracapacitor power source,” IEEE Trans-
[113] V. Yaramasu, B. Wu, P. C. Sen, S. Kouro, and M. Narimani, “High- actions on Power Electronics, vol. 29, no. 3, pp. 1469–1479, 2014.
power wind energy conversion systems: State-of-the-art and emerging [135] Amin, R. T. Bambang, A. S. Rohman, C. J. Dronkers, R. Ortega, and
technologies,” Proceedings of the IEEE, vol. 103, no. 5, pp. 740–788, A. Sasongko, “Energy management of fuel cell/battery/supercapacitor
2015. hybrid power sources using model predictive control,” IEEE Transac-
[114] S. Muller, M. Deicke, and R. De Doncker, “Doubly fed induction tions on Industrial Informatics, vol. 10, no. 4, pp. 1992–2002, 2014.
generator systems for wind turbines,” IEEE Industry Applications [136] “A Self-Adaptive Inertia and Damping Combination Control of
Magazine, vol. 8, no. 3, pp. 26–33, 2002. VSG to Support Frequency Stability - IEEE Journals & Magazine.”
[115] S. J. Chapman, Electric Machinery Fundamentals. McGraw-Hill, [Online]. Available: https://ieeexplore.ieee.org/document/7727967
2012. [137] C. Jia, J. Cui, W. Qiao, and L. Qu, “Real-time model predictive control
[116] Z. Zhang, “On control of grid-tied back-to-back power converters for battery-supercapacitor hybrid energy storage systems using linear
and permanent magnet synchronous generator wind turbine systems,” parameter varying models,” IEEE Journal of Emerging and Selected
Dissertation, Technische Universität München, München, 2016. Topics in Power Electronics, pp. 1–1, 2021.
[117] V. N. R. Yaramasu, “Predictive control of multilevel converters for [138] X. Zhang, B. Wang, D. Gamage, and A. Ukil, “Model predictive and
megawatt wind energy conversion systems. thesis,” Dissertation, Ryer- iterative learning control based hybrid control method for hybrid energy
son University, 2014. storage system,” IEEE Transactions on Sustainable Energy, vol. 12,
[118] V. Yaramasu, B. Wu, and J. Chen, “Model-Predictive Control of no. 4, pp. 2146–2158, 2021.
Grid-Tied Four-Level Diode-Clamped Inverters for High-Power Wind [139] L. Wang, Y. Wang, C. Liu, D. Yang, and Z. Chen, “A power distri-
Energy Conversion Systems,” IEEE Transactions on Power Electronics, bution strategy for hybrid energy storage system using adaptive model
vol. 29, no. 6, pp. 2861–2873, 2014. predictive control,” IEEE Transactions on Power Electronics, vol. 35,
[119] Z. Zhang, J. Rodrı́guez, and R. Kennel, “Advanced control strategies no. 6, pp. 5897–5906, 2020.
for direct-drive pmsg wind turbine systems: Direct predictive torque [140] S. Chen, Q. Yang, J. Zhou, and X. Chen, “A model predictive control
control approaches,” CPSS Transactions on Power Electronics and method for hybrid energy storage systems,” CSEE Journal of Power
Applications, vol. 2, no. 3, pp. 217–225, 2017. and Energy Systems, vol. 7, no. 2, pp. 329–338, 2021.
[120] Y. Zhang, W. Xie, Z. Li, and Y. Zhang, “Model predictive direct [141] H. H. Abdeltawab and Y. A.-R. I. Mohamed, “Robust energy manage-
power control of a pwm rectifier with duty cycle optimization,” IEEE ment of a hybrid wind and flywheel energy storage system considering
Transactions on Power Electronics, vol. 28, no. 11, pp. 5343–5351, flywheel power losses minimization and grid-code constraints,” IEEE
2013. Transactions on Industrial Electronics, vol. 63, no. 7, pp. 4242–4254,
[121] Z. Zhang, H. Fang, and R. Kennel, “Fully fpga based direct model 2016.
predictive power control for grid-tied afes with improved performance,”
[142] Z. Zhang, O. Babayomi, Z. Li, and C. Hu, “Hybrid wind-solar micro-
in IECON 2015 - 41st Annual Conference of the IEEE Industrial
grid rapid frequency response distributed coordination control method
Electronics Society, 2015, pp. 003 881–003 886.
and system,” China patent CN112 769 149A, 2021.
[122] H. Fang, Z. Zhang, X. Feng, and R. Kennel, “Ripple-reduced model
predictive direct power control for active front-end power converters [143] K. Tytelmaier, O. Husev, O. Veligorskyi, and R. Yershov, “A review of
with extended switching vectors and time-optimised control,” IET non-isolated bidirectional dc-dc converters for energy storage systems,”
Power Electronics, vol. 9, pp. 1914–1923(9), July 2016. in 2016 II International Young Scientists Forum on Applied Physics and
Engineering (YSF), 2016, pp. 22–28.
[123] Y. Zhang and W. Xie, “Low complexity model predictive con-
trol—single vector-based approach,” IEEE Transactions on Power [144] M. Forouzesh, Y. P. Siwakoti, S. A. Gorji, F. Blaabjerg, and B. Lehman,
Electronics, vol. 29, no. 10, pp. 5532–5541, 2014. “Step-up dc–dc converters: A comprehensive review of voltage-
[124] A. J. Sguarezi Filho and E. R. Filho, “Model-based predictive control boosting techniques, topologies, and applications,” IEEE Transactions
applied to the doubly-fed induction generator direct power control,” on Power Electronics, vol. 32, no. 12, pp. 9143–9178, 2017.
IEEE Transactions on Sustainable Energy, vol. 3, no. 3, pp. 398–406, [145] H. Sartipizadeh, F. Harirchi, M. Babakmehr, and P. Dehghanian,
2012. “Robust model predictive control of dc-dc floating interleaved boost
[125] L. Xu and P. Cartwright, “Direct active and reactive power control converter with multiple uncertainties,” IEEE Transactions on Energy
of dfig for wind energy generation,” IEEE Transactions on Energy Conversion, vol. 36, no. 2, pp. 1403–1412, 2021.
Conversion, vol. 21, no. 3, pp. 750–758, 2006. [146] Y. Liang, Z. Liang, D. Zhao, Y. Huangfu, and L. Guo, “Model
[126] Y. Zhang, Z. Li, Y. Zhang, W. Xie, Z. Piao, and C. Hu, “Performance predictive control for interleaved dc-dc boost converter based on
improvement of direct power control of pwm rectifier with simple kalman compensation,” in 2018 IEEE International Power Electronics
calculation,” IEEE Transactions on Power Electronics, vol. 28, no. 7, and Application Conference and Exposition (PEAC), 2018, pp. 1–5.
pp. 3428–3437, 2013. [147] W. Zhan and W. Wang, “Observer-based adaptive model predictive
[127] V. Ambrozic, R. Fiser, and D. Nedeljkovic, “Direct current control- control for interleaved boost dc-dc converter,” in 2019 4th International
a new current regulation principle,” IEEE Transactions on Power Conference on Intelligent Green Building and Smart Grid (IGBSG),
Electronics, vol. 18, no. 1, pp. 495–503, 2003. 2019, pp. 452–455.
[128] R. S. Herrera and P. Salmeron, “Instantaneous reactive power theory: A [148] S. P. Engel, M. Stieneker, N. Soltau, S. Rabiee, H. Stagge, and R. W.
comparative evaluation of different formulations,” IEEE Transactions De Doncker, “Comparison of the modular multilevel dc converter and
on Power Delivery, vol. 22, no. 1, pp. 595–604, 2007. the dual-active bridge converter for power conversion in hvdc and mvdc
[129] Y. Zhang, W. Xie, Z. Li, and Y. Zhang, “Model predictive direct grids,” IEEE Transactions on Power Electronics, vol. 30, no. 1, pp.
power control of a pwm rectifier with duty cycle optimization,” IEEE 124–137, 2015.
Transactions on Power Electronics, vol. 28, pp. 5343–5351, 2013. [149] L. Chen, S. Shao, Q. Xiao, L. Tarisciotti, P. W. Wheeler, and
[130] X. Wang and D. Sun, “Three-vector-based low-complexity model T. Dragičević, “Model predictive control for dual-active-bridge con-
predictive direct power control strategy for doubly fed induction verters supplying pulsed power loads in naval dc micro-grids,” IEEE
generators,” IEEE Transactions on Power Electronics, vol. 32, no. 1, Transactions on Power Electronics, vol. 35, no. 2, pp. 1957–1966,
pp. 773–782, 2017. 2020.
[131] C. Charumit and V. Kinnares, “Discontinuous svpwm techniques of [150] L. Chen, L. Lin, S. Shao, F. Gao, Z. Wang, P. W. Wheeler, and
three-leg vsi-fed balanced two-phase loads for reduced switching losses T. Dragičević, “Moving discretized control set model-predictive control
and current ripple,” IEEE Transactions on Power Electronics, vol. 30, for dual-active bridge with the triple-phase shift,” IEEE Transactions
no. 4, pp. 2191–2204, 2015. on Power Electronics, vol. 35, no. 8, pp. 8624–8637, 2020.
[132] J. I. LEON, E. Dominguez, L. Wu, A. Marquez Alcaide, M. Reyes, [151] L. Tarisciotti, L. Chen, S. Shao, T. Dragicevic, P. Wheeler, and
and J. Liu, “Hybrid energy storage systems: Concepts, advantages, and P. Zanchetta, “Finite control set model predictive control for dual active
applications,” IEEE Industrial Electronics Magazine, vol. 15, no. 1, pp. bridge converter,” IEEE Transactions on Industry Applications, pp. 1–
74–88, 2021. 1, 2021.
 21

[152] E. Hossain, R. Perez, A. Nasiri, and S. Padmanaban, “A comprehen- tational burden and fixed switching spectrum for a multilevel inverter in
sive review on constant power loads compensation techniques,” IEEE a photovoltaic system,” IEEE Access, vol. 8, pp. 77 405–77 414, 2020.
Access, vol. 6, pp. 33 285–33 305, 2018. [171] P. Falkowski, A. Sikorski, and M. Malinowski, “Finite control set
[153] Q. Xu, Y. Yan, C. Zhang, T. Dragicevic, and F. Blaabjerg, “An model predictive control with floating virtual voltage vectors for grid-
offset-free composite model predictive control strategy for dc/dc buck connected voltage source converter,” IEEE Transactions on Power
converter feeding constant power loads,” IEEE Transactions on Power Electronics, vol. 36, no. 10, pp. 11 875–11 885, 2021.
Electronics, vol. 35, no. 5, pp. 5331–5342, 2020. [172] M. E. Zarei, D. Ramirez, M. Prodanovic, and G. Venkataramanan,
[154] O. Andrés-Martı́nez, A. Flores-Tlacuahuac, O. F. Ruiz-Martinez, and “Multivector model predictive power control for grid connected con-
J. C. Mayo-Maldonado, “Nonlinear model predictive stabilization of verters in renewable power plants,” IEEE Journal of Emerging and
dc–dc boost converters with constant power loads,” IEEE Journal of Selected Topics in Power Electronics, vol. 10, no. 2, pp. 1466–1478,
Emerging and Selected Topics in Power Electronics, vol. 9, no. 1, pp. 2022.
822–830, 2021. [173] B. Liu, H. Wang, Y. Yang, X. Zhang, and B. Guo, “Improved model
[155] L. Cheng, P. Acuna, R. P. Aguilera, J. Jiang, S. Wei, J. E. Fletcher, and predictive control for single-phase grid-tied inverter with virtual vectors
D. D. C. Lu, “Model predictive control for dc–dc boost converters with in the compacted solution-space,” IEEE Transactions on Industrial
reduced-prediction horizon and constant switching frequency,” IEEE Electronics, vol. 69, no. 9, pp. 9673–9678, 2022.
Transactions on Power Electronics, vol. 33, no. 10, pp. 9064–9075, [174] H. Mahmoudi, M. Aleenejad, and R. Ahmadi, “Modulated Model
2018. Predictive Control of Modular Multilevel Converters in VSC-HVDC
[156] S.-K. Kim, C. R. Park, J.-S. Kim, and Y. I. Lee, “A stabilizing model Systems,” IEEE Transactions on Power Delivery, vol. 33, no. 5, pp.
predictive controller for voltage regulation of a dc/dc boost converter,” 2115–2124, 2018.
IEEE Transactions on Control Systems Technology, vol. 22, no. 5, pp. [175] J. Wang, X. Liu, Q. Xiao, D. Zhou, H. Qiu, and Y. Tang, “Modulated
2016–2023, 2014. Model Predictive Control for Modular Multilevel Converters With
[157] A. Kirubakaran, S. Jain, and R. Nema, “A review on fuel cell tech- Easy Implementation and Enhanced Steady-State Performance,” IEEE
nologies and power electronic interface,” Renewable and Sustainable Transactions on Power Electronics, vol. 35, no. 9, pp. 9107–9118,
Energy Reviews, vol. 13, no. 9, pp. 2430–2440, 2009. 2020.
[158] M. Sagar Bhaskar, V. K. Ramachandaramurthy, S. Padmanaban, [176] Y. Jin, Q. Xiao, H. Jia, Y. Mu, Y. Ji, T. Dragi?evi?, R. Teodorescu, and
F. Blaabjerg, D. M. Ionel, M. Mitolo, and D. Almakhles, “Survey of F. Blaabjerg, “A Novel Sliding-Discrete-Control-Set Modulated Model
dc-dc non-isolated topologies for unidirectional power flow in fuel cell Predictive Control for Modular Multilevel Converter,” IEEE Access,
vehicles,” IEEE Access, vol. 8, pp. 178 130–178 166, 2020. vol. 9, pp. 10 316–10 327.
[159] P. K. Seth, B. M. Reddy, and P. Samuel, “Comparative analysis of [177] M. R. Chowdhury, S. Chowdhury, M. A. Rahman, and M. R. Islam,
application of power electronic converters in fuel cell hybrid electric “Advanced switching sequences based model-predictive control for
vehicles: A review,” in 2018 3rd IEEE International Conference on single-phase npc converters,” IEEE Transactions on Industrial Elec-
Recent Trends in Electronics, Information Communication Technology tronics, vol. 69, no. 4, pp. 3515–3526, 2022.
(RTEICT), 2018, pp. 1518–1524. [178] F. Donoso, A. Mora, R. Cárdenas, A. Angulo, D. Sáez, and M. Rivera,
[160] D. F. Pereira, F. da Costa Lopes, and E. H. Watanabe, “Neural “Finite-set model-predictive control strategies for a 3l-npc inverter
generalized predictive control for tracking maximum efficiency and operating with fixed switching frequency,” IEEE Transactions on
maximum power points of pem fuel cell stacks,” in IECON 2018 - Industrial Electronics, vol. 65, no. 5, pp. 3954–3965, 2018.
44th Annual Conference of the IEEE Industrial Electronics Society,
[179] Q. Xiao, Y. Jin, H. Jia, Y. Mu, Y. Ji, R. Teodorescu, and F. Blaabjerg,
2018, pp. 1878–1883.
“Modulated model predictive control for multilevel cascaded h-bridge
[161] M. B. Shadmand, R. S. Balog, and H. Abu-Rub, “Model predictive
converter-based static synchronous compensator,” IEEE Transactions
control of PV sources in a smart DC distribution system: Maximum
on Industrial Electronics, vol. 69, no. 2, pp. 1091–1102, 2022.
power point tracking and droop control,” IEEE Transactions on Energy
[180] R. O. Ramı́rez, C. R. Baier, F. Villarroel, J. R. Espinoza, J. Pou,
Conversion, vol. 29, no. 4, pp. 913–921, dec 2014.
[162] M. A. Elgendy, B. Zahawi, and D. J. Atkinson, “Operating charac- and J. Rodrı́guez, “A hybrid fcs-mpc with low and fixed switching
teristics of the p amp;o algorithm at high perturbation frequencies frequency without steady-state error applied to a grid-connected chb
inverter,” IEEE Access, vol. 8, pp. 223 637–223 651, 2020.
for standalone pv systems,” IEEE Transactions on Energy Conversion,
vol. 30, no. 1, pp. 189–198, 2015. [181] C. A. Rojas, M. Aguirre, S. Kouro, T. Geyer, and E. Gutierrez, “Leak-
[163] ——, “Assessment of the incremental conductance maximum power age current mitigation in photovoltaic string inverter using predictive
point tracking algorithm,” IEEE Transactions on Sustainable Energy, control with fixed average switching frequency,” IEEE Transactions on
vol. 4, no. 1, pp. 108–117, 2013. Industrial Electronics, vol. 64, no. 12, pp. 9344–9354, 2017.
[164] P. Karamanakos, M. Nahalparvari, and T. Geyer, “Fixed switching [182] S. Vazquez, P. Acuna, R. P. Aguilera, J. Pou, J. I. Leon, and L. G. Fran-
frequency direct model predictive control with continuous and dis- quelo, “Dc-link voltage-balancing strategy based on optimal switching
continuous modulation for grid-tied converters with lcl filters,” IEEE sequence model predictive control for single-phase h-npc converters,”
Transactions on Control Systems Technology, vol. 29, no. 4, pp. 1503– IEEE Transactions on Industrial Electronics, vol. 67, no. 9, pp. 7410–
1518, 2021. 7420, 2020.
[165] IEEE, “Ieee recommended practice and requirements for harmonic [183] A. Mora, R. Cárdenas-Dobson, R. P. Aguilera, A. Angulo, F. Donoso,
control in electric power systems,” IEEE Std 519-2014 (Revision of and J. Rodriguez, “Computationally Efficient Cascaded Optimal
IEEE Std 519-1992), pp. 1–29, 2014. Switching Sequence MPC for Grid-Connected Three-Level NPC Con-
[166] R. Guzman, L. G. de Vicuña, A. Camacho, J. Miret, and J. M. Rey, verters,” IEEE Transactions on Power Electronics, vol. 34, no. 12, pp.
“Receding-horizon model-predictive control for a three-phase vsi with 12 464–12 475, dec 2019.
an lcl filter,” IEEE Transactions on Industrial Electronics, vol. 66, [184] A. Mora, R. Cardenas, R. P. Aguilera, A. Angulo, P. Lezana, and D. D.-
no. 9, pp. 6671–6680, 2019. C. Lu, “Predictive optimal switching sequence direct power control for
[167] C. R. D. Osório, D. A. Schuetz, G. G. Koch, F. Carnielutti, D. M. grid-tied 3l-npc converters,” IEEE Transactions on Industrial Electron-
Lima, L. A. M. Jr, V. F. Montagner, and H. Pinheiro, “Modulated model ics, vol. 68, no. 9, pp. 8561–8571, 2021.
predictive control applied to lcl-filtered grid-tied inverters: A convex [185] R. Pérez-Ibacache, A. L. Cedeño, C. A. Silva, G. Carvajal, J. C.
optimization approach,” IEEE Open Journal of Industry Applications, Agüero, and A. Yazdani, “Decentralized model-based predictive control
vol. 2, pp. 366–377, 2021. for der units integration in ac microgrids subject to operational and
[168] C. S. Lim, S. S. Lee, I. U. Nutkani, X. Kong, and H. H. Goh, “Near- safety constraints,” IEEE Transactions on Power Delivery, vol. 36,
optimal mpc algorithm for actively damped grid-connected pwm-vscs no. 4, pp. 2479–2489, 2021.
with lcl filters,” IEEE Transactions on Industrial Electronics, vol. 67, [186] Y. Shan, J. Hu, K. W. Chan, and S. Islam, “A unified model pre-
no. 6, pp. 4578–4589, 2020. dictive voltage and current control for microgrids with distributed
[169] R. Barzegarkhoo, S. A. Khan, Y. P. Siwakoti, R. P. Aguilera, S. S. Lee, fuzzy cooperative secondary control,” IEEE Transactions on Industrial
and M. N. H. Khan, “Implementation and analysis of a novel switched- Informatics, vol. 17, no. 12, pp. 8024–8034, 2021.
boost common-ground five-level inverter modulated with model predic- [187] S. Zhao, F. Blaabjerg, and H. Wang, “An overview of artificial
tive control strategy,” IEEE Journal of Emerging and Selected Topics intelligence applications for power electronics,” IEEE Transactions on
in Power Electronics, vol. 10, no. 1, pp. 731–744, 2022. Power Electronics, vol. 36, pp. 4633–4658, 2021.
[170] J. J. Silva, J. R. Espinoza, J. A. Rohten, E. S. Pulido, F. A. Villarroel, [188] M. Novak and T. Dragicevic, “Supervised imitation learning of finite-
M. A. Torres, and M. A. Reyes, “Mpc algorithm with reduced compu- set model predictive control systems for power electronics,” IEEE
 22

Transactions on Industrial Electronics, vol. 68, no. 2, pp. 1717–1723, [208] M. S. M. Gardezi and A. Hasan, “Machine learning based adaptive
2021. prediction horizon in finite control set model predictive control,” IEEE
[189] I. S. Mohamed, S. Rovetta, T. D. Do, T. Dragicević, and A. A. Z. Access, vol. 6, pp. 32 392–32 400, 2018.
Diab, “A neural-network-based model predictive control of three-phase [209] Z. Zhang, W. Tian, W. Xiong, and R. Kennel, “Predictive torque control
inverter with an output lc filter,” IEEE Access, vol. 7, pp. 124 737– of induction machines fed by 3L-NPC converters with online weighting
124 749, 2019. factor adjustment using Fuzzy Logic,” in 2017 IEEE Transportation
[190] S. Wang, T. Dragicevic, G. F. Gontijo, S. K. Chaudhary, and R. Teodor- Electrification Conference and Expo (ITEC), 2017, pp. 84–89.
escu, “Machine learning emulation of model predictive control for [210] F. Villarroel, J. R. Espinoza, C. A. Rojas, J. Rodriguez, M. Rivera, and
modular multilevel converters,” IEEE Transactions on Industrial Elec- D. Sbarbaro, “Multiobjective Switching State Selector for Finite-States
tronics, vol. 68, no. 11, pp. 11 628–11 634, 2021. Model Predictive Control Based on Fuzzy Decision Making in a Matrix
[191] S. Wang, T. Dragicevic, Y. Gao, and R. Teodorescu, “Neural network Converter,” IEEE Transactions on Industrial Electronics, vol. 60, no. 2,
based model predictive controllers for modular multilevel converters,” pp. 589–599, 2013.
IEEE Transactions on Energy Conversion, vol. 36, no. 2, pp. 1562– [211] H. Mahmoudi, M. Aleenejad, P. Moamaei, and R. Ahmadi, “Fuzzy ad-
1571, 2021. justment of weighting factor in model predictive control of permanent
[192] D. Wang, Z. J. Shen, X. Yin, S. Tang, X. Liu, C. Zhang, J. Wang, magnet synchronous machines using current membership functions,”
J. Rodriguez, and M. Norambuena, “Model predictive control using in 2016 IEEE Power and Energy Conference at Illinois (PECI), 2016,
artificial neural network for power converters,” IEEE Transactions on pp. 1–5.
Industrial Electronics, vol. 69, no. 4, pp. 3689–3699, 2022. [212] S. Wang, P. Dehghanian, M. Alhazmi, and M. Nazemi, “Advanced
[193] S. Lucia, D. Navarro, B. Karg, H. Sarnago, and s. Lucı́a, “Deep control solutions for enhanced resilience of modern power-electronic-
learning-based model predictive control for resonant power converters,” interfaced distribution systems,” Journal of Modern Power Systems and
IEEE Transactions on Industrial Informatics, vol. 17, no. 1, pp. 409– Clean Energy, vol. 7, no. 4, pp. 716–730, 2019.
420, 2021. [213] X. Zhou, F. Tang, P. C. Loh, X. Jin, and W. Cao, “Four-leg converters
[194] T. Dragicevic and M. Novak, “Weighting Factor Design in Model with improved common current sharing and selective voltage-quality
Predictive Control of Power Electronic Converters: An Artificial Neural enhancement for islanded microgrids,” IEEE Transactions on Power
Network Approach,” IEEE Transactions on Industrial Electronics, Delivery, vol. 31, pp. 522–531, 2016.
vol. 66, no. 11, pp. 8870–8880, 2019. [214] C. A. Rojas, J. R. Rodriguez, S. Kouro, and F. Villarroel, “Multiobjec-
[195] D. Zhao, K. Shen, L. Chen, Z. Wang, W. Liu, T. Yang, and P. Wheeler, tive Fuzzy-Decision-Making Predictive Torque Control for an Induction
“Improved active damping stabilization of dab converter interfaced Motor Drive,” IEEE Transactions on Power Electronics, vol. 32, no. 8,
aircraft dc microgrids using neural network-based model predictive pp. 6245–6260, 2017.
control,” IEEE Transactions on Transportation Electrification, pp. 1–1, [215] M. J. Lesani, H. Mahmoudi, M. Ebrahim, S. Varzali, and
2021. D. Arab khaburi, “Predictive torque control of induction motor based
[196] M. Novak, H. Xie, T. Dragicevic, F. Wang, J. Rodriguez, and F. Blaab- on improved fuzzy control method,” in 2013 13th Iranian Conference
jerg, “Optimal Cost Function Parameter Design in Predictive Torque on Fuzzy Systems (IFSC), 2013, pp. 1–5.
Control (PTC) Using Artificial Neural Networks (ANN),” IEEE Trans- [216] H. Mahmoudi, M. j. Lesani, and D. Arab khabouri, “Online fuzzy
actions on Industrial Electronics, pp. 1–1, 2020. tuning of weighting factor in model predictive control of PMSM,” in
[197] J. He, L. Xing, and C. Wen, “Weighting factors’ real-time updating 2013 13th Iranian Conference on Fuzzy Systems (IFSC), 2013, pp. 1–5.
for finite control set model predictive control of power converters via [217] M. A. Khan, A. Haque, V. S. B. Kurukuru, and S. Mekhilef, “Ad-
reinforcement learning,” in 2021 IEEE 16th Conference on Industrial vanced control strategy with voltage sag classification for single-phase
Electronics and Applications (ICIEA), 2021, pp. 707–712. grid-connected photovoltaic system,” IEEE Journal of Emerging and
[198] S. Vazquez, D. Marino, E. Zafra, M. Valdes, J. J. Rodriguez-Andina, Selected Topics in Industrial Electronics, vol. 3, no. 2, pp. 258–269,
L. G. Franquelo, and M. Manic, “An artificial intelligence approach for 2022.
real-time tuning of weighting factors in fcs-mpc for power converters,” [218] X. Zhang, Z. Li, Z. Zhang, M. Zhang, H. Chen, and Z. Zhang, “Neural
IEEE Transactions on Industrial Electronics, pp. 1–1, 2021. network based open-circuit fault diagnosis for three-level neutral-
[199] O. Machado, P. Martı́n, F. J. Rodrı́guez, and E. J. Bueno, “A Neural point-clamped back-to-back converters,” in 2021 IEEE International
Network-Based Dynamic Cost Function for the Implementation of Conference on Predictive Control of Electrical Drives and Power
a Predictive Current Controller,” IEEE Transactions on Industrial Electronics (PRECEDE), 2021, pp. 748–752.
Informatics, vol. 13, no. 6, pp. 2946–2955, 2017. [219] D. Koraki and K. Strunz, “Wind and solar power integration in
[200] F. Wang, J. Li, Z. Li, D. Ke, J. Du, C. Garcia, and J. Rodriguez, “Design electricity markets and distribution networks through service-centric
of model predictive control weighting factors for pmsm using gaussian virtual power plants,” IEEE Transactions on Power Systems, vol. 33,
distribution based particle swarm optimization,” IEEE Transactions on no. 1, pp. 473–485, 2018.
Industrial Electronics, pp. 1–1, 2021. [220] W. Wang, P. Chen, D. Zeng, and J. Liu, “Electric vehicle fleet
[201] X. Liu, L. Qiu, W. Wu, J. Ma, Y. Fang, Z. Peng, and D. Wang, integration in a virtual power plant with large-scale wind power,” IEEE
“Predictor-based neural network finite-set predictive control for modu- Transactions on Industry Applications, vol. 56, no. 5, pp. 5924–5931,
lar multilevel converter,” IEEE Transactions on Industrial Electronics, 2020.
vol. 68, no. 11, pp. 11 621–11 627, 2021. [221] L. Yavuz, A. Önen, S. Muyeen, and I. Kamwa, “Transformation of
[202] M. R. Habibi, H. R. Baghaee, F. Blaabjerg, and T. Dragicevic, microgrid to virtual power plant – a comprehensive review,” IET
“Secure mpc/ann-based false data injection cyber-attack detection and Generation, Transmission & Distribution, vol. 13, no. 11, pp. 1994–
mitigation in dc microgrids,” IEEE Systems Journal, pp. 1–12, 2021. 2005, 2019.
[203] X. Liu, L. Qiu, W. Wu, J. Ma, Y. Fang, Z. Peng, and D. Wang, “Event- [222] A. Parisio, E. Rikos, and L. Glielmo, “A model predictive control
triggered neural-predictor-based fcs-mpc for mmc,” IEEE Transactions approach to microgrid operation optimization,” IEEE Transactions on
on Industrial Electronics, vol. 69, no. 6, pp. 6433–6440, 2022. Control Systems Technology, vol. 22, no. 5, pp. 1813–1827, 2014.
[204] O. Babayomi, Z. Zhang, Y. Li, and R. Kennel, “Adaptive predictive [223] G. Mantovani, G. T. Costanzo, M. Marinelli, and L. Ferrarini, “Ex-
control with neuro-fuzzy parameter estimation for microgrid grid- perimental validation of energy resources integration in microgrids via
forming converters,” Sustainability, vol. 13, no. 13, 2021. distributed predictive control,” IEEE Transactions on Energy Conver-
[205] S. Sahoo, H. Wang, and F. Blaabjerg, “On the explainability of black sion, vol. 29, no. 4, pp. 1018–1025, 2014.
box data-driven controllers for power electronic converters,” in 2021 [224] A. Bolzoni, A. Parisio, R. Todd, and A. J. Forsyth, “Optimal virtual
IEEE Energy Conversion Congress and Exposition (ECCE), 2021, pp. power plant management for multiple grid support services,” IEEE
1366–1372. Transactions on Energy Conversion, vol. 36, no. 2, pp. 1479–1490,
[206] M. Ishaq and M. H. Afzal, “Supervised machine learning based 2021.
artificial neural network approach for the control of matrix converter,” [225] B. Zhou, X. Liu, Y. Cao, C. Li, C. Y. Chung, and K. W. Chan, “Optimal
in 2020 First International Conference of Smart Systems and Emerging scheduling of virtual power plant with battery degradation cost,” IET
Technologies (SMARTTECH), 2020, pp. 191–196. Generation, Transmission & Distribution, vol. 10, no. 3, pp. 712–725,
[207] M. Baker, H. Althuwaini, and M. B. Shadmand, “Resilient model based 2016.
predictive control scheme inspired by artificial intelligence methods [226] J. Kim, E. Muljadi, V. Gevorgian, M. Mohanpurkar, Y. Luo, R. Hov-
for grid-interactive inverters,” in 2021 6th IEEE Workshop on the sapian, and V. Koritarov, “Capability-coordinated frequency control
Electronic Grid (eGRID), 2021, pp. 01–06. scheme of a virtual power plant with renewable energy sources,” IET
 23

Generation, Transmission & Distribution, vol. 13, no. 16, pp. 3642– [238] Y. Du, W. Pei, N. Chen, X. Ge, and H. Xiao, “Real-time microgrid
3648, 2019. economic dispatch based on model predictive control strategy,” Journal
[227] D. Zhao, H. Wang, J. Huang, and X. Lin, “Virtual energy storage of Modern Power Systems and Clean Energy, vol. 5, no. 5, pp. 787–
sharing and capacity allocation,” in 2020 IEEE Power Energy Society 796, 2017.
General Meeting (PESGM), 2020, pp. 1–1. [239] O. Stanojev, S. Member, U. Markovic, S. Member, and P. Aristidou,
[228] X. Zhu, J. Yang, Y. Liu, C. Liu, B. Miao, and L. Chen, “Optimal “MPC-Based Fast Frequency Control of Voltage Source Converters in
scheduling method for a regional integrated energy system considering Low-Inertia Power Systems,” no. September, 2019.
joint virtual energy storage,” IEEE Access, vol. 7, pp. 138 260–138 272,
2019. [240] R. Heydari, M. Savaghebi, and F. Blaabjerg, “Fast frequency control
[229] Y. Wang, M. Yu, and Y. Li, “Self-adaptive inertia control of DC mi- of low-inertia hybrid grid utilizing extended virtual synchronous ma-
crogrid based on fast predictive converter regulation,” IET Renewable chine,” in 2020 11th Power Electronics, Drive Systems, and Technolo-
Power Generation, vol. 11, no. 8, pp. 1295–1303, 2017. gies Conference (PEDSTC). IEEE, 2020, pp. 1–5.
[230] J. Jongudomkarn, J. Liu, and T. Ise, “Virtual Synchronous Generator [241] Z. Yi, X. Zhao, D. Shi, J. Duan, Y. Xiang, and Z. Wang, “Accurate
Control with reliable Fault Ride-Through ability: A solution based on Power Sharing and Synthetic Inertia Control for DC Building Micro-
Finite-Set Model Predictive Control,” IEEE Journal of Emerging and grids with Guaranteed Performance,” IEEE Access, vol. 7, pp. 63 698–
Selected Topics in Power Electronics, no. September, pp. 1–1, 2019. 63 708, 2019.
[231] O. Babayomi, Z. Li, and Z. Zhang, “Distributed secondary frequency
[242] R. Heydari, Y. Khayat, M. Naderi, A. Anvari-Moghaddam, T. Dragice-
and voltage control of parallel-connected vscs in microgrids: A predic-
vic, and F. Blaabjerg, “A Decentralized Adaptive Control Method for
tive vsg-based solution,” CPSS Transactions on Power Electronics and
Frequency Regulation and Power Sharing in Autonomous Microgrids,”
Applications, vol. 5, pp. 342–351, 2020.
IEEE International Symposium on Industrial Electronics, vol. 2019-
[232] U. Tamrakar, T. M. Hansen, R. Tonkoski, and D. A. Copp, “Model
June, pp. 2427–2432, 2019.
Predictive Frequency Control of Low Inertia Microgrids,” in IEEE
International Symposium on Industrial Electronics, vol. 2019-June. [243] K. Oureilidis, K.-N. Malamaki, K. Gallos, A. Tsitsimelis, C. Dikaiakos,
Institute of Electrical and Electronics Engineers Inc., jun 2019, pp. S. Gkavanoudis, M. Cvetkovic, J. M. Mauricio, J. M. Maza Ortega,
2111–2116. J. L. M. Ramos, G. Papaioannou, and C. Demoulias, “Ancillary services
[233] K. Liu, T. Liu, Z. Tang, and D. J. Hill, “Distributed mpc-based market design in distribution networks: Review and identification of
frequency control in networked microgrids with voltage constraints,” barriers,” Energies, vol. 13, no. 4, 2020.
IEEE Transactions on Smart Grid, vol. 10, no. 6, pp. 6343–6354, 2019. [244] A. Kaushal and D. Van Hertem, “An overview of ancillary services and
[234] L. Meng, J. Zafar, S. K. Khadem, A. Collinson, K. C. Murchie, hvdc systems in european context,” Energies, vol. 12, no. 18, 2019.
F. Coffele, and G. M. Burt, “Fast frequency response from energy
storage systems—a review of grid standards, projects and technical [245] X. Zhang, L. Zhang, and Y. Zhang, “Model predictive current control
issues,” IEEE Transactions on Smart Grid, vol. 11, no. 2, pp. 1566– for pmsm drives with parameter robustness improvement,” IEEE Trans-
1581, 2020. actions on Power Electronics, vol. 34, no. 2, pp. 1645–1657, 2019.
[235] D.-T. Hoang, S. Olaru, A. Iovine, J. Maeght, P. Panciatici, and M. Ruiz, model,” IEEE Transactions on Industrial Electronics, vol. 68, pp. 993–
“Predictive control for zonal congestion management of a transmission 1003, 2021.
network,” in 2021 29th Mediterranean Conference on Control and [247] O. Babayomi, Y. Zhang, Y. Wang, Z. Li, Z. Zhang et al., “A com-
Automation (MED), 2021, pp. 220–225. parative study on weighting factor design techniques for the model
[236] I. Kalogeropoulos and H. Sarimveis, “Predictive control algorithms for predictive control of power electronics and energy conversion systems,”
congestion management in electric power distribution grids,” Applied 2021.
Mathematical Modelling, vol. 77, pp. 635–651, 2020. [248] P. Mancarella, “Mes (multi-energy systems): An overview of concepts
[237] M. A. Velasquez, J. Barreiro-Gomez, N. Quijano, A. I. Cadena, and and evaluation models,” Energy, vol. 65, pp. 1–17, 2014.
M. Shahidehpour, “Intra-hour microgrid economic dispatch based on
model predictive control,” IEEE Transactions on Smart Grid, vol. 11, [249] Z. Li, L. Wu, Y. Xu, S. Moazeni, and Z. Tang, “Multi-stage real-
no. 3, pp. 1968–1979, 2020. time operation of a multi-energy microgrid with electrical and thermal
[246] Y. Zhang, J. Jin, and L. Huang, “Model-free predictive current control energy storage assets: A data-driven mpc-adp approach,” IEEE Trans-
of pmsm drives based on extended state observer using ultralocal actions on Smart Grid, vol. 13, no. 1, pp. 213–226, 2022.

View publication stats