International Journal of Electrical and Computer Engineering (IJECE) Vol.
No.
December 2024, pp.
ISSN: 2088-8708.
DOI: 10.
11591/ijece.
A combined control method of supply harmonic current and source harmonic voltage for series hybrid active power filter Chau Minh Thuyen.
Hoai Thuong Nguyen.
Phan Thi Bich Thao Faculty of Electrical Engineering Technology.
Industrial University of Ho Chi Minh City.
Ho Chi Minh City.
Vietnam
Article Info
ABSTRACT
Article history:
The series hybrid active power filter (SHAPF) is known as a very effective harmonic filtering model in power systems.
Typically.
SHAPF is controlled by a control method based either on the harmonic voltage of the load or on the supply harmonic current.
However, the above two methods have the disadvantage of requiring the control coefficient much be large enough, which easily causes system instability.
Therefore, this paper presents a new control method for SHAPF.
It is a combined control method of the supply harmonic current and the source harmonic voltage.
The advantage of the proposed method is the ability to reduce the total harmonic distortion of the supply current and voltage applied to the load with a control coefficient that is not too large.
A fuzzy-proportional integral controller is designed for the proposed control method to reduce the compensation error in steady state under variable load conditions.
Mathematical models and simulation results have demonstrated the effectiveness of the proposed method in reducing the total harmonic distortion of the supply current, voltage applied to the load and minimize the compensation error at steady state.
Received Mar 11, 2024 Revised Jul 12, 2024 Accepted Jul 17, 2024 Keywords:
Active power filter Combined control method Hybrid active power filter Passive power filter Proportional integral controller This is an open access article under the CC BY-SA license.
Corresponding Author:
Chau Minh Thuyen Faculty of Electrical Engineering Technology.
Industrial University of Ho Chi Minh City 12 Nguyen Van Bao Road.
Go Vap District.
Ho Chi Minh City 700000.
Vietnam Email: chauminhthuyen@iuh.
INTRODUCTION
Today, due to the characteristics of manufacturing industries, electrical equipment is required to operate in many different modes, with most loads being controlled by power electronic devices.
As a result, the loads generate current harmonic components, which distort and shift phase compared to the source From there, a shunt active power filter (Shunt APF) was born to filter current harmonic components .
However, harmonic components are not only generated by power electronic loads, they are also generated by sources.
The harmonics generated by the source will impose a harmonic voltage on the load, which is very dangerous.
Therefore, the series active power filter .
eries APF) was born as a necessity .
Ae.
However, the series APF is only capable of filtering harmonic components from the source, but it is not capable of filtering harmonic components generated from the load.
For this reason, series hybrid active power filter (SHAPF) was born.
It is a combination of the series APF and passive power filters (PPF.
, .
PPFs have the function of filtering load current harmonics while the series APF filters voltage harmonics emitted from the source.
Therefore, research on SHAPF has practical significance, it contributes to improve the power quality of the power system.
Published research on SHAPF focusing on the design of PPFs for SHAPF is carried out in .
Ae.
Research on control strategies for SHAPF based on the load harmonic voltage and supply harmonic current is cited in .
Ae.
Many studies on control for SHAPF have also been carried out such as dead-beat control Journal homepage: http://ijece.
ISSN: 2088-8708
, sliding control .
, fuzzy sliding control .
, fuzzy-neural control .
, .
, neural network control .
, .
To increase the adaptability of the controller when the load changes and reduce the compensation error at steady state, some studies have hybridized controllers together such as proportional integral (PI)neural controller, fuzzy-neural controller .
Ae.
Stability assessment of the control algorithm using Laplace transform in the frequency domain is given in .
In .
Ae.
use a neural controller to online adjust the parameters of the fuzzy controller's membership function to reduce output error every time the load However, this analysis is only performed for the control strategy based on the harmonic voltage on the load.
In summary, previous studies on SHAPF only used control methods based on the supply harmonic current or harmonic voltage on the load.
With the control method based on the supply harmonic current, the control coefficient must be large enough, and the voltage applied to the load contains the harmonic component of the source voltage.
With the control method based on the harmonic voltage on the load, the control coefficient must be large enough to eliminate the harmonic component of the voltage on the load but cannot eliminate the harmonic component in the supply current.
Thus, the disadvantage of the above two methods is that the control coefficient must be large enough, which can easily cause system instability.
From the above analysis, this article proposes a new control method for SHAPF.
It is a control method based on the supply harmonic current and the source harmonic voltage.
The advantage of this method is that it can reduce the harmonic content in the supply current and the harmonic voltage applied to the load with a control coefficient not too large.
The harmonic voltage applied to the load will not be affected by the harmonic component of the source voltage.
This can increase the stability of the system and improve the harmonic filtering efficiency of SHAPF.
However, the actual load will vary.
Therefore, to respond well to load changes, a fuzzy-proportional integral (Fuzzy-PI) controller is designed for SHAPF with the proposed control method.
The structure of the article includes five sections.
An overview of SHAPF is presented in section 1.
Structure and control of SHAPF are introduced in section 2.
Fuzzy-PI controller design is given in section 3.
Simulation results and discussion are performed in section 4.
Finally, the research results are summarized in STRUCTURE AND CONTROL OF SHAPF The structure of a SHAPF is shown in Figure 1.
SHAPF consists of two main parts: PPFs and the active part .
nverter, output filter and transforme.
PPFs are designed to filter harmonic components emitted by the load, while the active part is designed to compensate for the remainder.
The active part can be viewed as an impedance.
It has a value of zero at the fundamental frequency and equal to infinity at the harmonic It prevents harmonic components from the source into the load.
U Ca U Cb U Cc Inverter U dc PPFs Figure 1.
The structure of a SHAPF In which: ycOyc and ycsyc are the voltage and impedance of the source.
The nonlinear load is modelled by two three-phase uncontrolled bridge rectifiers with loads R.
C, and RL.
PPFs are passive power filters designed to suppress the high harmonics of the nonlinear load.
ycOyayca , ycOyayca , and ycOyayca are the compensation voltages from the inverter.
ya0 ya0 is output filter of the inverter and ycOyccyca is power supply voltage for the Int J Elec & Comp Eng.
Vol.
No.
December 2024: 6057-6065
Int J Elec & Comp Eng
ISSN: 2088-8708
SHAPF consists of two main parts: PPFs and the active part .
nverter, output filter and transforme.
PPFs are designed to filter harmonic components emitted by the load, while the active part is designed to compensate for the remainder.
The active part can be viewed as an impedance.
It has a value of zero at the fundamental frequency and equal to infinity at the harmonic frequency.
It prevents harmonic components from the source into the load.
The single-phase equivalent circuit in the harmonic domain of SHAPF is shown in Figure 2.
To reduce harmonic components from the nonlinear load and source, the compensation voltage from the active part of APF can be based on the supply harmonic current or the load harmonic Therefore, ycOyca can be considered as a dependent voltage source.
From Figure 2, control methods for SHAPF can be listed as follows:
I sh
Z sh
U Lh
I Fh
U sh
PPF
I Lh Figure 2.
The single-phase equivalent circuit in the harmonic domain of SHAPF Control method ycOya = yayaycEa .
From Figure 2, yaycEa and ycOyaEa can be calculated as .
yaycEa = ycOycEa ycsycEycEyaEa yayaEa ycsycEa ya ycsycEycEyaEa ycOyaEa = ycOycEa Oe .
a ycsycEa )yaycEa .
From .
we can see that: if ya is large enough, the supply harmonic current yaycEa will decrease to zero.
However, the voltage applied to the load will have harmonic components of sources ycOycEa and ycOycEa Oe .
a ycsycEa )yaycEa will not decrease but increase.
On the other hand, if ya is too large, it can easily cause system Control method ycOya = yaycOyaEa .
From Figure 2, yaycEa and ycOyaEa can be calculated as .
yaycEa = ycOycEa .
ycsycEycEyaEa yayaEa ycsycEa .
ycsycEycEyaEa ycOyaEa = ycOycEa OeycsycEayaycEa 1 ya .
From .
we can see that: if ya is large enough, ycOyaEa will decrease to zero and prevent harmonic components into load from the source.
However, this method cannot reduce harmonic components in the supply current.
In short, this method is only used to reduce harmonic components of the source voltage applied to the load but is not capable of reducing the harmonic content of the supply current.
Proposed control method From the above two methods, this paper proposes a new control method.
It is a combination of the supply harmonic current and the source harmonic voltage ycOya = ya1 yaycEa ya2 ycOycEa .
From Figure 2, yaycEa and ycOyaEa can be calculated as .
yaycEa = .
Oeya2 )ycOycEa ycsycEycEyaEa yayaEa ycsycEa ya1 ycsycEycEyaEa ycOyaEa = .
Oe ya2 )ycOycEa Oe .
a1 ycsycEa )yaycEa .
From .
we can see that: if we control them so that ya2 = 1 and ya1 are not too large .
hey do not need to be too large as in method ycOya = yayaycEa ), we can reduce yaycEa and ycOyaEa .
In particular, this method has the A combined control method of supply harmonic current and source harmonic A (Chau Minh Thuye.
A ISSN: 2088-8708 ability to eliminate the influence of source harmonic voltage ycOycEa on yaycEa and ycOyaEa , and it can avoid instability due to too large K control coefficient as in method ycOya = yayaycEa .
K1 and K2 are controllable coefficients.
The reference signals yaycEa , ycOyaEa and ycOycEa are determined from yayc , ycOya and ycOyc using the ip-iq method .
as shown in Figure 3.
The components yaycyca , yaycyca , yaycyca are converted from the abc rotation coordinate system to the stationary coordinate system dq0, then pass through a low-pass filter to block the high-order harmonic components and pass the fundamental components.
Finally, the dq0Ieabc inverse transform is used to obtain the fundamental frequency components yaycyceyca , yaycyceyca , and yaycyceyca .
The reference harmonic components are obtained:
yaycEayca = yaycyca Oe yaycyceyca aycEayca = yaycyca Oe yaycyceyca yaycEayca = yaycyca Oe yaycyceyca The transformation between coordinates is shown through formulas .
ycaycuyc yuE ycaycuyc( yuE Oe 2yuU/.
ycaycuyc( yuE 2yuU/.
cnyc ] = .
cycnycu yuE ycycnycu( yuE Oe 2yuU/.
ycycnycu( yuE 2yuU/.
] .
cnycyca ] ycnycyca
1/21/21/2
ycaycuyc yuE ycycnycu yuE 1 cnycyceyca ] = .
caycuyc( yuE Oe 2yuU/.
ycycnycu( yuE Oe 2yuU/.
cnyc ] ycaycuyc( yuE 2yuU/.
ycycnycu( yuE 2yuU/.
1 ycn0 is the angle between the a and q axes or the angle between the a and d axes.
Components ycOyaEa and ycOycEa are also determined similarly.
I sa I sb I sc I sha I shb I shc I sfa I sfb I sfc Figure 3.
Determination of reference harmonic signal using the ip-iq method
DESIGN OF FUZZY-PI CONTROLLER FOR SHAPF
Because the load changes during control, the yaycE and yaya parameters of the PI controller need to be continuously adjusted according to the load change.
This paper designs a fuzzy regulator to adjust the parameters yaycE and yaya of the PI controller.
The control diagram for SHAPF using the Fuzzy-PI controller is shown in Figure 4.
U C Oe reference Fuzzy AEK P
Inverter
Output filter
AEK I
Figure 4.
Control diagram of SHAPF using Fuzzy-PI controller Int J Elec & Comp Eng.
Vol.
No.
December 2024: 6057-6065
Int J Elec & Comp Eng
ISSN: 2088-8708
The initial yaycE and yaya parameters are determined according to the Ziegler-Nichols method.
The outputs of the fuzzy regulator are yuuyaycE and yuuyaya .
They are adjusted based on the error between the reference signal and the real signal and the change of this error.
ya ycuyceyc = yaycE yuuyaycE { ycEycuyceyc = yaya yuuyaya The inputs and outputs of the fuzzy regulator are represented as seven membership functions: negative big (NB), negative medium (NM), negative small (NS), zero (Z.
, positive small (PS), positive medium (PM) and positive big (PB) as shown in Figures 5.
NB NM NS Z0 PS PM PB
NB NM NS Z0 PS PM PB
-10 -6 -3
e, de
-3 -1
AEK P .
AEK I .
Figure 5.
Membership functions of the fuzzy variable: .
membership functions of the e, de and .
membership functions of the yuuyaycE and yuuyaya Fuzzy rules are the core of the fuzzy adjustor and it is built based on the following principles:
Oe If .
is large, a large yuuyaycE value is needed.
If .
is small, a small yuuyaycE value is needed Oe If .
is approximately zero then the current yaya value is appropriate Oe If yce.
yccyce > 0 then a large yuuyaycE value is needed.
If yce.
yccyce < 0 then a small yuuyaycE value is needed Oe If .
are large, a large value of yuuyaycE and yuuyaya = 0 is needed The fuzzy rules are shown in Table 1.
The inference method uses the MAX-MIN method.
The centroid method is used for defuzzification.
Table 1.
Fuzzy rules of yuuyaycE and yuuyaya
yuuyaycE /yuuyaya yce
PB/ZO
PM/ZO
PM/ZO
PM/ZO
PS/ZO
ZO/ZO
ZO/ZO
PB/ZO
PM/ZO
PM/ZO
PS/ZO
PS/ZO
ZO/ZO
NS/ZO
PM/NB
PM/NM
PS/NS
ZO/NS
ZO/ZO
NS/PS
NB/NS
PM/NM
PS/NM
PS/NS
ZO/NM
NS/PS
NM/PM
NM/PM
PS/NM
PS/NS
ZO/ZO
NS/PS
NS/PS
NM/PM
NM/PB
PS/ZO
ZO/ZO
NS/ZO
NM/ZO
NM/ZO
NB/ZO
NB/ZO
ZO/ZO
ZO/ZO
NM/ZO
NM/ZO
NB/ZO
NB/ZO
NB/ZO
SIMULATION RESULTS AND DISCUSSION
To demonstrate the effectiveness of the proposed control method.
The simulation results are performed on a SHAPF system as shown in Figure 1.
The parameters of a SHAPF system are given in Table 2.
Table 2.
Parameters of a SHAPF system Parameters Source Source impedance Load Transformer ratio PPF: Consists of branches L and C connected in parallel Output filter Inverter DC voltage Value (Us-rms=220 V.
f=50 H.
(Upeak=25 V.
f=1000 H.
Rs = 0.
1 A.
Ls = 0.
2 mH R = 20 A.
L = 30 mH.
C = 500 F.
RL=50 A
L =16.
9 mH.
C = 600 F
L0 = 2 mH.
C0 = 60 F
A combined control method of supply harmonic current and source harmonic A (Chau Minh Thuye.
A ISSN: 2088-8708 The waveforms that need to be considered include: source voltage ycOycyca , load current yayayca , supply current yaycyca , voltage applied to load ycOyayca and compensation error of phase a.
At t=0.
3 s the load changes in the direction of increasing total harmonic distortion (THD).
Oe With control method ycOyca = yayaycEa using a PI controller.
Control coefficient K=50.
The simulation results of the waveforms with control method ycOyca = yayaycEa are shown in Figure 6.
Oe With control method ycOyca = yaycOyaEa .
PI controller is used.
Control coefficient K=50.
The simulation results of the waveforms with control method ycOyca = yaycOyaEa are shown in Figure 7.
Oe With control method ycOyca = ya1 yaycEa ya2 ycOycEa using a PI controller.
Control coefficient K=20.
The simulation results of the waveforms with control method ycOyca = ya1 yaycEa ya2 ycOycEa using a PI controller are shown in Figure 8.
Oe With control method ycOyca = ya1 yaycEa ya2 ycOycEa using Fuzzy-PI controller.
Control coefficient K=20.
The simulation results of the waveforms with control method ycOyca = ya1 yaycEa ya2 ycOycEa using a Fuzzy-PI controller are shown in Figure 9.
U sa I La Time .
I sa Time .
U La Time .
Time .
Figure 6.
Simulation results of the waveforms with control method ycOyca = yayaycEa U sa I La Time .
I sa Time .
U La Time .
Time .
Figure 7.
Simulation results of the waveforms with control method ycOyca = yaycOyaEa Int J Elec & Comp Eng.
Vol.
No.
December 2024: 6057-6065
Int J Elec & Comp Eng
ISSN: 2088-8708
U sa I La Time .
I sa Time .
U La Time .
Time .
Figure 8.
Simulation results of the waveforms with control method ycOyca = ya1 yaycEa ya2 ycOycEa using a PI controller U sa I La Time .
I sa Time .
U La Time .
Time .
Figure 9.
Simulation results of the waveforms with control method ycOyca = ya1 yaycEa ya2 ycOycEa using a Fuzzy-PI The summary table of simulation results with control methods is shown in Table 3.
From the results in Table 3, we can see that the control method ycOyca = yayaycEa with K=50 is capable of reducing THD of the supply current and voltage applied to load.
However, it is only effective when the control coefficient K is large enough.
Control method ycOyca = yaycOyaEa with K=50 is only capable of reducing THD of the voltage applied to the load.
However, it is only effective when the control coefficient K is large enough.
Control method ycOyca = ya1 yaycEa ya2 ycOycEa using a PI controller is capable of reducing THD of the supply current and the voltage applied to the load with a control coefficient K is not too large (K=.
, and especially it has the ability to eliminate the influence of the source voltage harmonic on the voltage applied to the load.
To further improve the efficiency of the proposed control method when the load changes suddenly, a fuzzy regulator is used to adjust the KP.
KI parameters of the PI controller.
The simulation results in Table 3 show that the proposed control method using a Fuzzy-PI controller is more effective than using a PI controller in reducing THD and compensation error at steady state.
A combined control method of supply harmonic current and source harmonic A (Chau Minh Thuye.
A ISSN: 2088-8708 Table 3.
Summary of simulation results with control methods Control methods ycOycyca ycOyca = yayaycEa ycOyca = yaycOyaEa ycOyca = ya1 yaycEa ya2 ycOycEa using a PI controller ycOyca = ya1 yaycEa ya2 ycOycEa using a Fuzzy-PI controller THD% yayayca yaycyca ycOyayca 0sy0.
ycOycyca
A15
very large A10
THD%
yayayca yaycyca ycOyayca 3sy0.
A15
very large A10
CONCLUSION
This paper has performed a control analysis for SHAPF and proposed a new control method for SHAPF, which is a combination of supply harmonic current and source harmonic voltage.
The proposed method is capable of reducing the harmonic contents of the supply current and the harmonic voltage applied to the load with a control coefficient is not too large.
This contributes to improving the stability of the A Fuzzy-PI controller is designed to improve the efficiency of the traditional PI controller in variable load situations.
The simulation results have demonstrated the effectiveness of the proposed method compared to the control method only based on the supply harmonic current or harmonic voltage of the load.
REFERENCES