0% found this document useful (0 votes)
0 views

Active Disturbance Rejection-Based Model Predictive Control to Improve Power Quality With Reduced Current Sensors in EV Charging Stations

The document discusses a proposed active disturbance rejection-based model predictive control (ADR-MPC) strategy aimed at enhancing power quality in electric vehicle charging stations (EVCS) while reducing the number of required current sensors. This approach simplifies control structures by compensating for local load current harmonics and reactive power using only grid-side current measurements, thereby improving grid stability and resilience. Simulation results validate the effectiveness of the proposed method in achieving enhanced power quality performance and robust disturbance rejection.

Uploaded by

Bruno Lahm
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
0% found this document useful (0 votes)
0 views

Active Disturbance Rejection-Based Model Predictive Control to Improve Power Quality With Reduced Current Sensors in EV Charging Stations

The document discusses a proposed active disturbance rejection-based model predictive control (ADR-MPC) strategy aimed at enhancing power quality in electric vehicle charging stations (EVCS) while reducing the number of required current sensors. This approach simplifies control structures by compensating for local load current harmonics and reactive power using only grid-side current measurements, thereby improving grid stability and resilience. Simulation results validate the effectiveness of the proposed method in achieving enhanced power quality performance and robust disturbance rejection.

Uploaded by

Bruno Lahm
Copyright
© © All Rights Reserved
We take content rights seriously. If you suspect this is your content, claim it here.
Available Formats
Download as PDF, TXT or read online on Scribd
You are on page 1/ 5

2025 16th Annual Power Electronics & Drives: Systems and Technologies Conference (PEDSTC) 4-6 February 2025,

University of Tabriz, Tabriz, Iran

Active Disturbance Rejection-Based Model


Predictive Control to Improve Power Quality with
2025 16th Power Electronics, Drive Systems, and Technologies Conference (PEDSTC) | 979-8-3315-3394-6/25/$31.00 ©2025 IEEE | DOI: 10.1109/PEDSTC65486.2025.10912064

Reduced Current Sensors in EV Charging Stations


Mahdi S. Mousavi∗ , S. Alireza Davari∗ , Majid Rahmani∗ , Behnam Nikmaram∗ ,
Mokhtar Aly† , Freddy Flores-Bahamonde‡ , Jose Rodriguez†
∗ Department of Electrical Engineering, Shahid Rajaee Teacher Training University, Tehran, Iran
† Faculty
of Engineering, Architecture and Design, Universidad San Sebastian, Santiago, Chile
‡ Energy Transformation Center, Engineering Faculty, Universidad Andres Bello, Santiago, Chile

Email: [email protected]

Abstract—This paper proposes an active disturbance rejection The EVCS provides opportunities to improve power quality
(ADR)–based model predictive control (MPC) strategy to enhance in the grid by applying advanced control techniques. In other
power quality in electric vehicle charging stations (EVCS) with words, an EVCS as a power quality improvement system can
a reduced number of current sensors. The presented approach
compensates for local load current harmonics and reactive power enhance grid stability and resilience without needing a specific
without the need for harmonic extraction, simplifying the control power converter to address these issues [6]. Therefore, EVCS
structure. By relying on current measurements only at the grid alongside supplying EV charging demand can enhance power
side, the method eliminates the need for additional sensors in quality by mitigating current harmonics and reactive power
the load or converter of the EVCS. Load current dynamics are [6], [7]. For instance, in [6], the EVCS improved the power
considered as disturbance terms, then are rejected using the ADR
technique. The proposed converter system simultaneously im- quality in a distorted and unbalanced grid using the model
proves the grid’s power quality and fulfills its primary function of predictive control (MPC) approach. A similar power structure
charging EVs. Simulation results validate the effectiveness of the has been utilized in [7] to achieve an EVCS with a fast
approach, demonstrating enhanced power quality performance harmonic rejection ability applying the MPC with a two-vector
and robust disturbance rejection. optimization approach. In [8], the instantaneous power theory
Index Terms—Model predictive control, EV, charging station,
has been employed to extract the load harmonics generated by
disturbance, power quality, reduced sensor EV chargers and local loads. Then a hysteresis current control
was applied to control the EVCS aiming for power quality
I. I NTRODUCTION improvement of the grid and supplying the DC loads including
battery chargers. In [9], an EVCS has been constructed using a
The growing interest in electric vehicles (EVs) and the multiphase system with multi-winding transformers to improve
expansion of EV charging stations (EVCS) within power the grid power quality without requiring a high switching
distribution networks are reshaping modern power systems frequency.
and increasing nonlinear and variable loads. Thus, power In these studies, enabling the grid support for the EVCS
quality issues, such as voltage sags, harmonic distortions, and requires additional control loops for extracting and mitigating
reactive power are significantly increased in these grids. These load harmonics. Moreover, extra current measurements are
problems can result in grid instability, reduced efficiency, needed at the load side for these approaches.
and potential equipment failure. Therefore, it is essential to To solve the mentioned problems, this paper focuses on
mitigate power quality problems in the grid effectively [1]. a control strategy that leverages active disturbance rejection
Generally, the power quality problems are solved in the grid (ADR) within the model predictive control (MPC) framework
using systems such as static compensators (STATCOMs) [2] to optimize EVCS performance. By employing only grid-side
and active power filters (APFs) [3]. A distribution STATCOM measurements and eliminating the need for local load current
has been employed in [2] to mitigate the harmonics and sensors, the approach simplifies implementation while main-
reactive power of an EVCS. The performance of the EVCS taining effective harmonic and reactive power compensation.
has been improved in [4] by employing an additional dynamic Simulation results demonstrate that this method achieves stable
voltage restorer (DVR) to enhance the power quality of the DC link voltage regulation for EV charging and sustains grid
grid. An APF has been utilized in [5] to mitigate the current current quality despite load variations.
harmonics of the grid, so the EVCS works properly in charging
the EV batteries. It should be noted that in these studies II. MPC OF EV C HARGING S TATIONS
the EVCS infrastructure becomes complicated by employing Fig. 1 presents the schematic of the system under study,
additional power converter for power quality improvement. which includes a grid, a local bus, and a three-phase converter

979-8-3315-3394-6/25/$31.00 ©2025 IEEE


Authorized licensed use limited to: UNIVERSIDADE FEDERAL DE SANTA MARIA. Downloaded on May 22,2025 at 19:01:54 UTC from IEEE Xplore. Restrictions apply.
Local Bus 3ph. Converter BCU PCC EVCS
ig ,abc i f ,abc
Grid Line Filter
Grid Lf , Rf
BCU Vdc
ABC iL ,abc
BCU
αβ

Load Harmonic Extraction


Nonlinear load ABC
Local SVPWM
3ph.
BCU:
iL Loads αβ
Motor
Battery Charger Unit PQ Calc. if v*inv
(6) v pcc v pcc Prediction  Vdc*
Local Loads
(5)
PL QL 
Fig. 1. Schematic of the EVCS in a grid with local loads.
Harmonics i*h i*f
Calc. (7)
 u PI
p.u
functioning as the EVCS. On the DC side of the converter, Model Predictive Current Control
multiple battery charger units (BCUs) are connected to serve
EVs. The grid also supplies local loads connected to the bus, Fig. 2. Block diagram of the MPC for EVCS with power
which consist of motor loads and nonlinear loads. quality enhancing feature.
The MPC with power quality enhancing ability for the
EVCS is implemented in two parts. The first part ensures the sample ik+1 equal to the reference current i∗f . For an EVCS,
f
stable operation of the EVCS by controlling both the DC link the reference current is obtained from the DC link voltage
voltage and the AC side currents. These controls are essential controller. Considering a proportional-integral (PI) controller
for maintaining the functionality of the charging station under for the DC link voltage regulation, the reference current can
varying load and grid conditions. The second part introduces be defined as follows:
an optional feature aimed at enhancing grid power quality by
kI
mitigating harmonic currents and compensating for reactive i∗f = (kP + )(Vdc ∗
−Vdc )vp.u
pcc , (4)
power. Both parts are discussed in detail in the following s
subsections. where kP and kI are the proportional and integral gains of
the PI controller, Vdc ∗ represents the reference value of the DC
p.u
A. Control of DC Link Voltage and AC Side Currents link voltage, and vpcc denotes the per-unit voltage of the PCC.
p.u
The DC link voltage and AC side currents are the controlled Notably, multiplying the PI controller’s output by vpcc ensures
variables in the EVCS. Fig. 2 illustrates the block diagram of that the reference current is in phase with the grid voltage,
the conventional MPC. The system’s dynamic behavior can be thereby enabling unity power factor operation.
described using the differential equations of the filter utilized Finally, putting the reference current instead of ik+1f leads
at the converter’s output: the following deadbeat voltage reference:
i∗f − ikf
Lf
di f
+ R f i f + vinv = vpcc , (1) v∗inv = vkpcc − R f ikf − L f . (5)
dt Ts
  The obtained reference voltage v∗inv is applied to the
where i f = i f α i f β is the filter’s current vector; L f and R f
are the inductance and resistance of the filter; the voltage at converter through a modulation technique, such as space
the by vpcc = vector pulse width modulation (SVPWM), to generate the
 point of common

coupling (PCC) is depicted
 required switching signals. The following subsection extends
vpcc,α vpcc,β , and vinv = vinv,α vinv,β is the inverter’s
voltage vector defined as follows: the control strategy by incorporating features for power quality
improvement, including load harmonic mitigation and reactive
2   power compensation.
vinv,α + jvinv,β = Vdc sa + sb a + sc a2 , (2)
3 B. Enabling Power Quality Support of EVCS
where Vdc is the DC-link voltage, sx , x ∈ {a, b, c} depicts the The power quality enhancing feature of the EVCS is im-
switching state of the appointed leg, and a = e j2π/3 . When the plemented by utilizing the instantaneous PQ theory [10]. This
upper switch is on, sx = 1, and when it is off, sx = 0. approach calculates the active and reactive power components
The MPC predicts future system behavior using the discrete- of the local loads connected to the PCC based on the measured
time equivalent of the model (1), which is formulated as load currents and voltages as follows:
follows:
ik+1
f − ikf PL = vpcc,α iLα + vpcc,β iLβ ,
vkinv = vkpcc − R f ikf − L f , (3) (6)
Ts QL = vpcc,β iLα − vpcc,α iLβ ,
where Ts denotes the sampling time and k represents the where iLα and iLβ are the load current components.
sampling instant. In the MPC with a deadbeat solution, the The calculated load power PL consists of both fundamental
reference voltage is achieved by setting the future current and harmonic components, i.e., PL = P̄L + P̃L where P̄L and

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DE SANTA MARIA. Downloaded on May 22,2025 at 19:01:54 UTC from IEEE Xplore. Restrictions apply.
P̃L denote the fundamental and harmonic components, respec- PCC EVCS
tively. The harmonic component needs to be mitigated on the Grid
ig ,abc L f , R f i f ,abc
grid side. Therefore, the harmonic term P̃L should be extracted Vdc
iL ,abc
using a high-pass filter. Additionally, the entire load reactive
power QL is included in the compensation process to achieve ABC Local
Vdc*
a unity power factor on the grid side. These components are αβ Loads SVPWM 
utilized in the subsequent equation to determine the reference ˆi v*inv PI
harmonic current required for mitigating distortions in the grid v pcc ig g i*g
Prediction
current [10]. ESO (12) D̂ u
(14)

∗     p.u
ihα 1 vpcc,α vpcc,β P̃L
i∗h = ∗ = 2 . (7) Proposed Predictive Current Control
ihβ vpcc,α + vpcc,β
2 −v pcc,β vpcc,α QL
Fig. 3. Block diagram of the proposed ADR-based MPC for
The obtained reference harmonic current is combined with the EVCS.
the reference current derived in (4), forming the total reference
current of the EVCS with enabled power quality improvement.
where e denotes the error between the measured and esti-
kI
i∗f = i∗h + (kP +∗
)(Vdc −Vdc )vp.u
pcc . (8) mated voltages; superscript “ ˆ ” depicts the estimated values;
s η1 and η2 are the observer’s gain. These gains can be tuned
In Fig. 2, the control components enabling power quality according to the bandwidth design method in [12].
improvement for the EVCS are highlighted in orange, while To apply the MPC, (10) is transformed into the discrete
the essential parts of the controller are highlighted in blue. domain as follows:
g − ig
ik+1 k
III. P ROPOSED ADR-BASED MPC FOR EVCS k
vkinv = vkpcc − L f + L f D̂ . (13)
Ts
In this section, we propose an ADR-based MPC method for
It is worth noting that the estimated total disturbance from
the EVCS with the ability to improve the power quality with-
(12) is utilized in the above equation. The MPC with the
out measuring load currents and extracting their harmonics.
deadbeat solution is achieved by replacing the future sample
Therefore, the proposed method simplifies the control strategy
of the grid current with the reference grid current as follows:
and reduces the number of required current sensors, making
it more cost-effective and easier to implement in practical i∗g − ikg k
applications. To develop this approach, the dynamic equation v∗inv = vkpcc − L f + L f D̂ , (14)
Ts
(1) is rewritten by applying ig = i f + iL where ig denotes the
grid current vector. Thus, (1) is updated as follows: where i∗g represents the grid current reference vector. To
ensure a harmonic-free grid current and maintain unity power
d(ig − iL ) factor, the grid current must consist solely of the fundamental
Lf + R f (ig − iL ) + vinv = vpcc . (9)
dt component of the load ifund
L and the fundamental component
Using the ADR theory [11], the above equation can be of the EVCS ifund
f . In other words, the grid current reference
reformulated as: is defined as follows:
dig vpcc − vinv i∗g = ifund
L + ifund
f . (15)
= + D, (10)
dt Lf The fundamental component of the EVCS current is derived
where D represents the total disturbance, which consists of the from the DC link voltage PI controller as defined in (4). On the
internal dynamics of the system and the external disturbance other hand, the fundamental component of the load current can
modeled by d in the following equation: be considered a proportion of the voltage at PCC. Therefore,
the grid current reference can be defined as:
diL R f Rf
D= − ig + iL + d. (11) kI
dt Lf Lf i∗g = kL vp.u
pcc + (kP +

)(Vdc −Vdc )vp.u
pcc , (16)
s
This paper utilizes an extended state observer (ESO) to where kL denotes the proportional gain for determining the
estimate the total disturbance. The ESO is designed based on load fundamental current, which cannot be precisely obtained
the following set of equations: without a load current sensor. In the proposed method, setting
⎧ kL = 0 leads to properly controlling the grid current in the

⎪ e = ig − îg


steady state. This approach is justified by the assumption that

⎨ dîg vpcc − vinv theoretically, the EVCS supplies the load current, while the
= + D̂ + η1 e , (12)
dt Lf grid provides power to the EVCS. In other words, the load is



⎪ indirectly powered through the EVCS. The block diagram of

⎩ dD̂ = η2 e the proposed method has been presented in Fig. 3.
dt

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DE SANTA MARIA. Downloaded on May 22,2025 at 19:01:54 UTC from IEEE Xplore. Restrictions apply.
IV. R ESULTS AND D ISCUSSION DC-link Voltage [100 V/div]
The proposed ADR-based MPC method was validated
through simulations in MATLAB/Simulink. System specifi- 700 V
cations for the simulation setup are summarized in Table I.
It should be noted that the nonlinear load is modeled with a
three-phase full-bridge diode rectifier with an RC load on its
Load Current [40 A/div]
DC side.
Fig. 4 presents the EVCS initializing performance where
the DC link is controlled without applying BCU loads. The
waveforms in this figure are the DC link voltage, local load
current, and grid current, respectively from top to bottom.
The DC link voltage is effectively controlled at the desired Grid Current [40 A/div]
reference (700 V), which is required for stable EV charging.
The local loads produce distorted currents, indicating harmonic
and reactive components. Despite this, the proposed control
method compensates for these distortions on the grid side
indirectly, without the need to sense the local load currents.
The resulting grid current exhibits minimal harmonic distor-
tion, demonstrating the effective compensation achieved by the Fig. 4. Initializing the EVCS with power quality improving
control approach. ability.
Fig. 5 illustrates the system’s performance when a BCU
load is applied to the DC side, representing the activation of DC-link Voltage [100 V/div]
charging. The DC link voltage, DC load current (a BCU with 700 V
10 A load), local load current (AC side), and grid current
are shown from top to bottom. When the DC load is applied,
the DC link voltage initially experiences a drop but quickly DC LoadCurrent [5 A/div]
recovers to the reference. An increase in grid current occurs as
expected due to the charging current demand, while the har- Load Current [40 A/div]
monic and reactive power compensation remains unaffected,
confirming that the control scheme can simultaneously support
both charging and power quality improvement on the grid side.
In Fig. 6, the DC load is disconnected, simulating a scenario
where charging is disabled. Under this condition, the DC link Grid Current [40 A/div]
voltage momentarily rises before settling back to the reference
level, and the DC load current falls to zero. As expected, the
grid current decreases accordingly. The compensation of local
load harmonics and reactive power continues uninterrupted,
illustrating the method’s robustness in maintaining power
quality independent of the DC load state.
Finally, Fig. 7 provides a steady-state comparison to high- Fig. 5. Performance of the proposed ADR-based MPC when
light harmonic compensation and power factor correction the DC load of the EVCS is applied suddenly.
(PFC). The upper plot shows the grid voltage alongside the
local load currents, where a noticeable phase lag and distortion
in the local load current reflect the reactive and harmonic characteristics of the load. The lower plot displays the grid
voltage and grid current in phase alignment, indicating suc-
TABLE I cessful PFC and reactive power compensation. This alignment
S YSTEM SPECIFICATIONS emphasizes the method’s effectiveness in improving grid-side
power quality, in the presence of distorted local loads.
Name Symbol Value
Rated line voltage VL−L 380 V V. C ONCLUSION
Frequency f 50 Hz
DC link voltage reference VDC 700 V This paper presented an ADR-based MPC strategy to en-
Filter inductance Lf 1 mH hance grid power quality using the EVCS with reduced current
Filter resistance Rf 0.1 Ω
Nonlinear load capacitance CL 400 μF sensing requirements. By employing a disturbance rejection
Nonlinear load resistance RL 40 Ω approach, the proposed method enables effective compensation
Motor load S 30 kVA of harmonics and reactive power without the need to sense
Power factor of the Motor load cosφ 0.8
local load currents, relying only on grid-side measurements.

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DE SANTA MARIA. Downloaded on May 22,2025 at 19:01:54 UTC from IEEE Xplore. Restrictions apply.
700 V
ACKNOWLEDGMENTS
DC-link Voltage [100 V/div] J. Rodriguez acknowledges the support of ANID through
projects FB0008, 1210208, and 1221293.
DC LoadCurrent [5 A/div] This work was also supported by ANID, Chile
FONDECYT Iniciacion 11230430, and SERC-Chile
Load Current [40 A/div] ANID/FONDAP/1523A0006.
R EFERENCES
[1] A. M. Dos Santos Alonso, L. De Oro Arenas, J. P. Bonaldo,
J. De A. olímpio Filho, F. P. Marafão, and H. K. M. Paredes, “Power
quality improvement in commercial and industrial sites: An integrated
Grid Current [40 A/div] approach mitigating power oscillations,” IEEE Access, vol. 12, pp.
50 872–50 884, 2024.
[2] C. Balasundar, C. K. Sundarabalan, S. N. Santhanam, J. Sharma,
and J. M. Guerrero, “Mixed step size normalized least mean fourth
algorithm of dstatcom integrated electric vehicle charging station,” IEEE
Transactions on Industrial Informatics, vol. 19, no. 6, pp. 7583–7591,
2023.
[3] B. Long, T. Cao, X. Qi, D. Shen, J. M. Guerrero, J. Rodríguez, and K. T.
Chong, “Noninteger lexicographic-optimization-based sequential model-
predictive fault-tolerant control of t-type shunt active power filter,” IEEE
Fig. 6. Performance of the proposed ADR-based MPC when Transactions on Power Electronics, vol. 37, no. 6, pp. 7169–7184, 2022.
the DC load of the EVCS is removed. [4] S. M. Rao, R. J. Satputaley, R. K. Keshri, V. B. Borghate, and
A. Khergade, “Improving dc-bus voltage regulation in grid fed ev fast
charging station using mli based dvr,” IEEE Transactions on Power
Delivery, vol. 39, no. 4, pp. 2530–2540, 2024.
Voltage [300 V/div] [5] A. Srivastava and S. Saravanan, “Harmonic mitigation using optimal
active power filter for the improvement of power quality for a electric
vehicle changing station,” e-Prime - Advances in Electrical Engineering,
Electronics and Energy, vol. 8, p. 100527, 2024. [Online]. Available:
https://www.sciencedirect.com/science/article/pii/S2772671124001098
Load Current [100 A/div] [6] H. T. Nguyen, A. S. A. Sumaiti, K. A. Hosani, K. A. Jaafari, Y.-J.
Byon, J. Y. Alsawalhi, and M. S. E. Moursi, “Enhanced performance of
Voltage [300 V/div] charging stations via converter control under unbalanced and harmonic
distorted grids,” IEEE Transactions on Power Delivery, vol. 36, no. 6,
pp. 3964–3976, 2021.
[7] H. T. Nguyen, K. A. Hosani, A. S. Al-Sumaiti, T. H. Nguyen, K. A. A.
Jaafari, J. Y. Alsawalhi, and M. S. El Moursi, “Fast harmonic rejecting
Grid Current [100 A/div] control design to enable active support of charging stations to micro-
grids under distortion,” IEEE Transactions on Transportation Electrifi-
cation, vol. 9, no. 3, pp. 4132–4146, 2023.
[8] A. Verma and B. Singh, “Aff-sogi-drc control of renewable energy based
Fig. 7. Steady state performance of the proposed method in grid interactive charging station for ev with power quality improvement,”
controlling the grid current with PFC ability. IEEE Transactions on Industry Applications, vol. 57, no. 1, pp. 588–597,
2021.
[9] M. R. Khalid, M. S. Alam, M. Krishnamurthy, E. A. Al-Ammar,
H. Alrajhi, and M. S. J. Asghar, “A multiphase ac–dc converter with
Simulation results confirm that the system maintains a stable improved power quality for ev charging station,” IEEE Transactions on
DC link voltage for reliable EV charging while simultaneously Transportation Electrification, vol. 8, no. 1, pp. 909–924, 2022.
improving power quality by mitigating the effects of distorted [10] H. Akagi, E. H. Watanabe, and M. Aredes, Instantaneous power theory
and applications to power conditioning. John Wiley & Sons, 2017.
local loads. Even under varying load conditions, the method [11] J. Han, “From pid to active disturbance rejection control,” IEEE trans-
demonstrated robust performance, providing both charging actions on Industrial Electronics, vol. 56, no. 3, pp. 900–906, 2009.
functionality and PFC without compromising grid-side com- [12] M. S. Mousavi, S. Alireza Davari, V. Nekoukar, C. Garcia, and J. Ro-
driguez, “Model-free finite set predictive voltage control of induction
pensation. This approach represents a practical solution for motor,” in 2021 12th Power Electronics, Drive Systems, and Technolo-
enhancing grid support capabilities in EVCS. gies Conference (PEDSTC), 2021, pp. 1–5.

Authorized licensed use limited to: UNIVERSIDADE FEDERAL DE SANTA MARIA. Downloaded on May 22,2025 at 19:01:54 UTC from IEEE Xplore. Restrictions apply.

You might also like