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

net/publication/282505426

PLL-less Robust Active and Reactive Power Controller for Single Phase Grid-
connected Inverter with LCL Filter

Conference Paper · March 2015

DOI: 10.1109/APEC.2015.7104647

CITATIONS READS

30 370

2 authors:

Xiao Li Robert Balog

Texas A&M University Texas A&M University
62 PUBLICATIONS 953 CITATIONS 218 PUBLICATIONS 6,251 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Xiao Li on 04 October 2015.

The user has requested enhancement of the downloaded file.

 PLL-less Robust Active and Reactive Power
Controller for Single Phase Grid-connected Inverter
with LCL Filter
Xiao Li1, Student Member, IEEE and Robert S. Balog2, Senior Member, IEEE
Renewable Energy & Advanced Power Electronics Research Laboratory
Department of Electrical & Computer Engineering, Texas A&M University
College Station, Texas, 77843, USA
1
[email protected], [email protected]

Abstract—Grid connected inverters, especially those used in function is embedded into the power controller. In this way,
DG (Distributed Generation) system, are expected with the PLL unit is not required, so the complexity for designing
decoupled and flexible power control ability. Many existing the system can be decreased. Meanwhile, the proposed system
power control methods are based on synchronous reference keeps good performance of decoupled real and reactive power
frame transformation, which requires the phase angle control with excellent grid current quality under different
information provided by Phase-Locked Loop (PLL). This creates conditions of grid disturbance.
additional drawbacks of the system and makes the system
complex. This paper presents a controller for single-phase grid Inverter system with LCL filter is a high order system and
connected inverter system with LCL filter in stationary reference has high resonant peak value. To overcome the promising
frame. And PLL is avoided. Compared with traditional power resonance problem, active damping technique with double-
control scheme for single phase grid-connected inverter loop current control is applied into the controller design of this
controller, this control scheme has faster response, higher system. Traditional grid current controller for single phase
reliability and decoupled real and reactive power control ability. inverter with LCL filter is based on PI controller, which is not
Moreover, the proposed controller has good robustness under able to follow a sinusoidal reference without steady state error
different distorted grid conditions. Since the controller is totally due to the dynamics of the integral term. And its current
implemented in stationary reference frame, the control algorithm tracking performance will be worse under non-ideal grid
comes to be simpler and easier. The designed current controller, condition. Though this can be resolved by feedforward control
combining the active damping technique, helps restrain the high
of grid voltage to some extent, this can increase the
resonant peak value. Compared with traditional PI based current
controller, the designed current controller, which is based on the
complexity of system. In the designed inner current loop,
Proportional & Resonant (PR) controller, provides the system Proportional & Resonant (PR) controller is used to provide the
with high gain in wide range of frequency to track given system with rapid response and high gain in wide range of
reference value fast and output excellent grid current with low frequency.
THD. In this paper, we will analyze and evaluate the proposed
controller for single-phase grid connected inverter with LCL
Keywords—single-phase inverter; grid-connected; Phase-Lock
filter. Reliable performance under various grid distortion
Loop; LCL filter; active damping; decoupled power control; PR
controller; grid distortion
cases, high-quality output current and decoupled real and
reactive power control abilities of inverter system are achieved
I. INTRODUCTION in this way. Both simulation and experiment results show the
effectiveness of the proposed controller.
Many power control strategies for the single-phase inverter
have been proposed [1-4]. Generally, the synchronous II. GRID-CONNECTED INVERTER
reference transformation and current component decoupling Fig.1 shows the topology of an LCL-filter-based grid-
algorithm are used in control process, which complexes the connected inverter. The DC bus could be supplied by
control algorithm [5-7]. And this requires the amplitude and renewable energy resources, such as photovoltaic, wind
phase angle information of the AC mains voltage captured by generator, storage battery, super capacitor, etc. The inverter is
the PLL. However, the PLL module, as a nonlinear part, used to convert electricity energy from DC to AC and could be
usually degrades the output performance of system. And much bi-directional. The LCL filter restrains the high-frequency
effort is needed to design the parameters in PLL to get the harmonics.
desired performance, which increases the complexity and cost
of design. The relationships between Vs, Vg and I2 in figure 1 can be
expressed in (1). Figure 2 shows the bode plot of the open-
Moreover, due to the nature of weak grid, the grid voltage loop transfer function from Vo to I2. Obviously, there exists a
may have various disturbances such as sag, swell, frequency rather high resonant peak in the LCL filter, which is undesired
distortion, and contamination with voltage harmonics. The for the control system design.
disturbances in grid can degrade the performance of PLL
largely [8]. In the proposed controller, the synchronization
 values, 𝐼𝑝 and 𝐼𝑞 ,can be obtained. Then 𝐼𝑝 and 𝐼𝑞 are
S1 S2 multiplied with two unit sinusoidal signals coming from the
L1 L2 I2 second order general integrator. In this way, two new signals,
Cdc
Vs 
𝐼𝑝′ and 𝐼𝑞′ , will be achieved and contain both angle and
S3 S4 Vo C Vg

amplitude information, from which we can get the reference
signal of grid current. In the proposed controller, the inner
Fig. 1. Power stage of single-phase LCL-filter-based
loop of the system is the current loop, whose main function is
1 L1Cs 2  1  to improve the dynamic performance and ensure that the
I 2  VO  Vg
L1 L2Cs  ( L1  L2 ) s
3
L1 L2Cs 3  ( L1  L2 ) s output current can track the reference well with low THD.
Capacitor current feedback is also added into the current loop
Bode Diagram to overcome the resonance problem introduced by LCL filter.
150
A. Current Controller
Magnitude (dB)

100
Figure 4 shows the block diagram of current controller.
50

0
Vc
1
-50
-90
Cs
I ref
Phase (deg)

-135 
   1 
 1  
-180 GPR K kPWM L1 s L2s
 Vg
-225
IL2 Ic 
-270
10
1 2
10 10
3
10
4 Fig. 4. Simplified block diagram of double loop control of
Frequency (Hz) current controller for LCL filter
Fig. 2. Bode plots of the transfer function from Vo to I2 The current reference value can be derived based on figure
III. CONTROLLER IN STATIONARY REFERENCE FRAME 3 as
Figure 3 shows the structure of single-phase grid- I pVg   I qVg  
connected inverter system with the proposed controller. I ref  I p'  I q'  
| Vg |
L1 L2
Ig Here, PR controller is applied to serve as the grid current
Vdc C Vg loop compensator, which can provide infinite gain at the
resonance frequency ω0 to reach zero steady-state error. The
Peak value
vg  SOGI SOGI transfer function of PR controller is given in (4). The
g detection

| Vg | vg   ig   ig  corresponding bode plot is shown in figure 5.

Instantaneous
SPWM Ig  Power
Estimation
Pout Qout
' Bode Diagram
Ic I p Ip 
 P '
150
P  PI ref
 
Magnitude (dB)

 
PR   100

PI
 Qref'
'
I q
Iq 50

Fig.3. Control structure of proposed power control strategy

0
90
The grid-connected inverter is controlled with a multi-loop
controller composed of a power outer loop and a current inner 45
Phase (deg)

loop with active damping technique. The reactive power 0

reference command is typically provided by the grid operator -45

or user. This can be set to zero for unity, negative for leading, -90
0 1 2 3
and positive for the lagging power factor. The active power 10 10
Frequency (Hz)
10 10

reference is determined by the user or related with the

instantaneous output power value supplied in the system, Fig.5. Bode plot of ideal PR controller with Kp=1,KR=100
taking it as an example, for PV application, the active power 2K R s
reference should be the maximum power point determined Gs ( s)  K P  
from MPPT algorithm. s 2  02
In the proposed controller, the instantaneous real power The dual closed loop current controller with capacitor
and reactive power values are calculated based on the current inner-loop and grid current outer-loop aims at current
instantaneous power theory. By comparing them with accuracy and acceptable distortion. The open-loop transfer
corresponding reference values, two independent control function can be expressed as
 B1s 2  B2 s  B3 When Vg changes from 250-350V, the system performance
Go ( s)   does not change a lot.
A1s 5  A2 s 4  A3 s 3  A4 s 2  A5 s
Bode Diagram

 K P KK PWM 
50
Where B1 Vg=350
Vg=300

Magnitude (dB)
Vg=250

 B2  2KK PWM K R  0

 B3  K P KK PWM 02  -50

0

A1  L1L2C 

Phase (deg)
-45

-90

 A2  KK PWM L2C  -135

-180 -1
0 1 2 3
10 10 10 10 10

 A3  L1L2C  L1  L2 
2
0
Frequency (Hz)

Fig.7. Bode plot of whole system when Vg amplitude change and power
 A4  L2CKK PWM  2
0 
(controller parameters keep as :kp=0.02,kI=0.5)
This characteristic can not only improve the reliability
 A5  ( L1  L2 )02  performance of system under distorted grid conditions, but
also make it possible to apply the peak value detection method
Based on the pole placement theory, the desired design into the system. The peak value detection method is applied to
parameters can be achieved. Here, to get the good dynamic get the maximum value of grid voltage, as shown in figure 3.
performance of system, we choose KP=3.3, KR=3091.7. The peak detection method can be implemented by taking the
detected peak value in one or two cycles as the maximum
B. Power Controller
value of grid voltage. Though this method may create error in
After designing the current loop, the block diagram of the detection result, the detection error will not affect much on
power loop can be represented as figure 6, taking the control the performance of the whole system as mentioned above.
delay time into consideration. In this way, the bandwidth of
outer loop should be less than that of inner loop. C. Second Order Generating Integrator
To create an orthogonal signal from an original single-
Qref I ref Ig Pout phase signal, different OSG techniques have been proposed in
 kP s  kI 1 k 2
Vg the literature. In the proposed controller, the second-order
s Tc s  1 s  k s   2
2
generalized integrator (SOGI)-based OSG is applied. This
Qout 
approach prevents harmonics/ noises from reaching the
controller and therefore is suitable for this study. The input-to-
Fig.6. Block diagram of power loop
output transfer function describing the dynamics of SOGI is
The open-loop transfer function of the power controller shown in (6) and (7), where w is the fundamental angular
can be expressed as frequency.

kP s  kI k 2 1 x ( s) ks
G0 ( s)   Vg  2    2 
s s  ks   Tc s  1
2
x( s) s  ks   2
Where, kp and kI is the parameters of PI controller; k and 𝜔 x ( s) k 2
are two parameters of SOGI module, which will be introduced  2 
in detail in the following part. Tc is the control period of the x( s) s  ks   2
whole system. It is very small in real implementation.
The response of SOGI to the grid distortion is shown in
For the power outer loop, the main functions are to track figure 8.
the given power reference with zero steady-state error and
In figure 8(a), ideal grid voltage (Vg=300V;fvg=60Hz) is
ensure the stability of the system. Considering these control
designed to be combined with 8% 5th ,8% 7th, 8% 11th
aims, the specific controller parameters can be determined
harmonics. And it sags by 20% at time 0.07s and resume to
with frequency response method.
original condition at time 0.17s. In figure 8(b), the frequency
Normally, Vg, the amplitude of grid voltage, will be a of ideal grid voltage changes from 60Hz to 58Hz at 0.07s. And
changeable variable in the distorted grid conditions. As can be it resume to original condition at time 0.17s.
seen in (5), Vg is also a variable that may affect the system
From the above results, it can be seen that the SOGI
performance. Here, due to the existence of SOGI module,
module have a good performance in various distortion
which offering a high gain in the working frequency, the
conditions, especially in grid harmonics distorted condition.
change of Vg will not affect the system’s performance heavily
in a certain frequency range. This can be proved in figure 7. With the orthogonal signals created by SOGI modules, the
instantaneous power values can be calculated based on 3) When the grid frequency step changes from 60 Hz to 58
instantaneous power theory. The corresponding calculation Hz at 0.08 s and then resumes to original condition at 0.15 s,
equations are shown in (8) and (9). the output of system can be seen in figure 11. In this case, the
proposed control method can enforce grid current tracking grid
1 frequency with constant output power even though grid
P  (Vg   I g   Vg    I g   )  frequency changes. And the THD of 𝐼𝑔 is as low as 1.24%
2
1
Q  (Vg    I g   Vg   I g   ) 
Vg

2 data1
data2
200
data3
0
Where, Vg−α , Vg−β are the derived orthogonal signals of grid
data4
data5
-200
voltage, and Ig−α , Ig−β are the derived orthogonal signals of 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
grid current. data9
data10 Ig Time (s)
data11
data12
5
SOGI performance with input of distorted grid SOGI--Input (Vg)
SOGI--Output1
300
SOGI--Output2
0
200

100 -5
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time (s)
0
P
-100
600

-200
400

200
-300
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22
Time (s) 0
Fig.8(a) Grid voltage with 10%THD harmonics sags by 20% at 0.07s 0.05 0.1 0.15 Time (s) 0.2
SOGI performance with input of distorted grid SOGI--Input(Vg)
SOGI--Output1 Q
300
SOGI--Output2 100
200

100 0

-100
-100 0.05 0.1 0.15 0.2
Time (s)
-200
Fig.9. Output of system with grid with harmonics
-300
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 0.22
Time (s)

Fig.8(b).Grid frequency steps from 60Hz to 58Hz at 0.07s Vg

 200

0
IV. SIMULINK RESULTS -200

The simulation model based on figure 1 is built in 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time (s)
MATLAB. The component parameters are listed in Table I. 5
Ig

Table I. MAIN PARAMETERS OF SIMULATION MODULE

0
Dc bus voltage 400 V Frequency of Grid voltage 60 Hz
L1 6 mH L2 1.2 mH -5
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
C 10 uF 𝐑𝐚𝐭𝐢𝐧𝐠 𝐨𝐟 𝐨𝐮𝐭𝐩𝐮𝐭 𝐏𝐨𝐰𝐞𝐫 500 W Time (s)

Amplitude of Grid 300 V Switching frequency 20 kHz

voltage P
600

A. Performance Under Various Distorted Grid Conditions 400

1) When grid has 4% 3th, 5th, 7th and 11th order harmonics 200

at 0.08 s and then resumes to ideal grid at 0.15 s, the output of 0

system can be seen in figure 9. The THD of 𝐼𝑔 is as low as 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
Time (s)
0.2
Q
2.49% under proposed control method. 100

2) When grid voltage (𝑉𝑔 = 300 𝑉) sags by 10% at time 0

0.08 s and then resumes to original condition at time 0.15 s,

the output of system can be seen in figure 10. In this case, the -100
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
proposed control method ensures the constant output power Time (s)
Fig.10. Output of system with grid voltage sag
and perfect grid current even though grid voltage sags, and the
THD of 𝐼𝑔 is as low as 1.01%.
 Vg
steady state condition, when the real power reference value is
given as 500 W and reactive power reference value is given as
200
0 VA. In this case, the output grid current tracks the reference
0
well with good waveform as shown in results.
-200

0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
The dynamic performance of the converter with the
Ig
Time (s) proposed scheme is presented in figure 13(b), where the real
5
power reference value step changes from 500 W to 200 W and
reactive power reference value changes from 300 VA to -300
0
VA. Both of the changes are set to happen at 0.1 s. These
-5 results show the decoupled power control of the proposed
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time (s)
control strategy. And the dynamic process is very fast.
P
600

400

200

0
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2 Vg
Time (s)
Q
200

100

0
ig
-100
0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0.2
Time (s)

Fig.11. Output of system with grid frequency distortion

B. Decoupled Real and Reactive Power Control & Dynamic

Performance Fig. 13(a). Steady-state performance
When the reference value of real power and reactive power
step changes simultaneously at 0.1s and 0.2s, the system
output performance and given reference values are shown in
figure 12.
Vg

Real Power Output Performance Reactive Power Output Performance

800 300

600
P-output
P-reference 200
Q-output
Q-reference
ig
100
400
0
(W)

(W)

200
-100
0
-200
Fig. 13(b). Dynamic performance
-200 -300

-400 -400
0 0.05 0.1 0.15 0.2 0.25 0.3 0 0.05 0.1 0.15 0.2 0.25 0.3
time (s) time (s) VI. CONCLUSIONS
Fig.12 Step response of real and reactive power (Bidirectional) In this paper, a simplified and robust controller for single-
phase grid connected inverter with LCL filter is proposed. The
synchronization function is directly combined in power loop to
As one may see from figure 12, the system keeps stable replace the function of PLL. A double-loop current controller
under different step change conditions and the system output with PR compensator and active damping function is
tracks the reference value well in less than 0.1 s, which means implemented to achieve better dynamic performance and
a good dynamic performance. The real and reactive can larger stability range in this paper. Reliable performance under
control separately. And control reference value of them can be various grid distortion cases, high-quality output current and
both negative and positive. decoupled real and reactive power control abilities of inverter
system are achieved in this way. The whole control system is
V. EXPERIMENTAL RESULTS implemented in stationary reference frame, which helps
To verify the above analysis, the proposed control strategy simplify the control algorithm.
is implemented with dSPACE1007 with parameters in Table
1.
Figure 13(a) demonstrates the system performance under
 [8] S.-K. Chung, “A Phase Tracking System for Three Phase Utility
VII. REFERENCES Interface Inverters,” IEEE Trans. Power Electron., vol. 15, no. 3, pp.
[1] R. I. Bojoi, L. R. Limongi, D. Roiu, and A. Tenconi, “Enhanced 431-438, May 2000.
power quality control strategy for single-phase inverters in distributed [9] Q. Zhang, et al., “Analysis and Design of a Digital Phase-Locked Loop
generation systems,” IEEE Trans. Power Electron., vol. 26, no.3, pp. for Single-Phase Grid-Connected Power Conversion Systems,” IEEE
798–806, Mar. 2011. Trans. Ind. Electron., vol.58,no.8, pp. 3581-3592, Aug. 2011.
[2] S. Xu, J. Wang, and J. Xu, “A current decoupling parallel control [10] M. Rubens, et al., “Comparison of Three Single-Phase PLL Algorithms
strategy of single-phaseInverterwith voltage and current dual closed- for UPS Applications,” IEEE. Trans. Ind. Electron., vol. 55, no. 8, pp.
loop feedback,” IEEE Trans. Ind. Electron., vol. 60, no. 4, pp. 1306– 2923-2931, Aug. 2008.
1313, Apr. 2013. [11] Liu W, Hao X, Yang X, et al. A multi-resonant sliding-mode controller
[3] P. Sun, C. Liu, J.-S. Lai, and C.-L. Chen, “Grid-tie control of cascade for single-phase grid-connected inverter with LCL-filter[C]//Applied
dualbuck inverter with wide-range power flow capability for renewable Power Electronics Conference and Exposition (APEC), 2013 Twenty-
energy applications,” IEEE Trans. Power Electron., vol. 27, no. 4, pp. Eighth Annual IEEE. IEEE, 2013: 2541-2546.
1839–1849, Apr. 2012. [12] Guofei T, Guochun X, Zhibo Z, et al. A control method with grid
[4] B. Bahrani., et al., “Vector control of single-phasevoltagesource disturbances suppression for a single-phase LCL-filter-based grid-
converters based on fictive-axis emulation,” IEEE Trans. Ind. Appl., connected inverter[C]//Applied Power Electronics Conference and
vol. 47, no.2, pp. 831– 840, Mar. 2011 Exposition (APEC), 2012 Twenty-Seventh Annual IEEE. IEEE, 2012:
[5] Shadmand M B , Balog R S, Abu Rub H, “Maximum Power Point 1489-1493.
Tracking using Model Predictive Control of a flyback converter for [13] Cao B, Chang L. A variable switching frequency algorithm to improve
photovoltaic applications,”Power and Energy Conference at Illinois the total efficiency of single-phase grid-connected
(PECI), 2014. IEEE, 2014: 1-5. inverters[C]//Applied Power Electronics Conference and Exposition
[6] Shadmand M B, Mosa M, Balog R S, et al, “An improved MPPT (APEC), 2013 Twenty-Eighth Annual IEEE. IEEE, 2013: 2310-2315.
technique for high gain DC-DC converter using model predictive [14] Khajehoddin S A, Karimi-Gharteman M, Bakhshai A, et al. High
control for photovoltaic applications,”Applied Power Electronics quality output current control for single phase grid-connected
Conference and Exposition (APEC), 2014 Twenty-Ninth Annual IEEE. inverters[C]//Applied Power Electronics Conference and Exposition
IEEE, 2014: 2993-2999. (APEC), 2014 Twenty-Ninth Annual IEEE. IEEE, 2014: 1807-1814.
[7] Q.-C. Zhong, et al., “Self-Synchronized Synchronverters: Inverters
Without a Dedicated Synchronization Unit,” IEEE Trans. Power
Electron., vol. 29, no. 2, pp. 617-630, Feb. 2014.

View publication stats