International Journal of Engineering Research

Volume No.5, Issue No.12, pp : 948-953

ISSN:2319-6890 (online),2347-5013(print)
1 December 2016

Comparative Analysis between PI and Fuzzy Controllers for Dc Voltage

Control of Unbalanced 4 Wire Shunt Active Power Filter
1

Ahmed A. Mansour , 2Mona M. Mamdouh , 3Sameh F.S. Eid

1,2
Electronic Research Institute (ERI), Cairo
3
October University for Modern sciences & Arts (MSA), Cairo
1
[email protected], [email protected] , [email protected]
Abstract-This research proposes a shunt active power filter
(SAPF) for harmonics mitigation in order to improve power
quality of the electrical network. The reference harmonic
current extraction technique used was based on DQ0 power
theory under unbalanced non-linear loads conditions for four
wire-installed systems. The DC-link voltage plays an important
role in SAPF design to guarantee both steady high voltage
level for harmonics current chopping and support the inverter
power losses. In order to maintain constant DC link voltage,
two different controllers are used; PI controller and fuzzy logic
controller. A comparative analysis has done between the two
controllers. Hysteresis PWM has used for current
conditioning.
Keywords- Power quality, Harmonics current, Shunt active
power filter, DQ0 extraction technique, PI controller, Fuzzy
logic controller, Hysteresis PWM-unbalanced loads.
I. Introduction
Power Quality is defined as Set of parameters defining the
properties of power quality as delivered to the user in normal
operating conditions in terms of continuity of supply and
characteristics of voltage (symmetry, frequency, magnitude and
waveform) in IEC. This means that a perfect power supply
would be one that is always available, always within voltage and
frequency tolerances and has a pure noise- free sinusoidal wave
shape. Just how much deviation from perfection can be tolerated
depends on the users application, the type of equipment
installed and his view of his requirements. IEEE defined power
quality disturbances into seven categories based on wave shape:
[1-5].
1. Transients
2. Waveform distortion
3. Interruptions
4. Voltage fluctuations
5. Sag / Under voltage
6. Frequency variations
7. Swell / Overvoltage
This has leaded to the proposal of more stringent requirements
regarding power quality and standards such IEEE-519, which
reflects
these
preoccupations.
IEEE
Std.1159-1995,
recommended Practice for Monitoring Electric Power Quality.
This standard covers recommended methods of measuring power
quality events as shown in Figure1.
Wave distortion is the most important problem in power quality
as it is a steady state deviation from an ideal sine wave. The most
general type of wave distortion is harmonics. In recent years,
with the rapid development of the power electronic technology,
various power electronic devices and nonlinear elements,
IJER@2016

especially of many kinds of rectifier and switcher, are applied

widely in power system.

Figure 1. Definitions of Events by IEEE Std.1159.

Harmonics have produced serious adverse effects in the
equipment of power generators and distribution system, and of
their control and protection, as well as the users and the
communication fields[6]. There are many problems caused by
harmonic currents such as overloading of neutrals, Overheating
of transformers, Skin effect and higher Losses[7]. The total
harmonic distortion (THD) of currents and voltages is shown in
equation (1) and (2).

Where; Ih,Vh are currents and voltages of a certain harmonic

order.There are two types of harmonics matigation methods. The
first one is passive filters. Its disadvantages is that the source
impedance, which is not accurately known and varies with the
system configuration, strongly influences filtering characteristics
of the shunt passive filter [8].The filter may be overloaded when
the harmonic current components increase. At a specific
frequency an anti resonance or parallel resonance occurs
between the source impedance and the shunt passive filter [9].
As the load changes, a passive filter may need to be
disconnected, converted or replaced by new one[10]. The second
harmonic mitigation method is shunt active power filter
(SAPF).To compensate the distorted currents,SAPF injects
currents equal but opposite to the harmonic components, thus

doi : 10.17950/ijer/v5s12/1216

Page 948

International Journal of Engineering Research

Volume No.5, Issue No.12, pp : 948-953

ISSN:2319-6890 (online),2347-5013(print)
1 December 2016

only the fundamental components flow in the point of common

coupling (PCC). Four wire SAPF is connected in parallel to the
loads, unbalanced and non-linear
loads as shown in
Figure2,which causes the supply currents to be completely
sinusoidal and balanced [11-12].

Figure 2. Basic shunt active power filter compensating system.

II. Reference current extraction technique based on (DQ0)
The performance of this filter depends on three parts:
1.
2.
3.

Reference current extraction circuit and control technique.

PWM inverter technique
The control of D.C capacitor voltage and the circuit
responsible on this control.
Reference current extraction technique is the heart of the SAPF
design. it is implemented in three stages. There are many
techniques used to extract the compensating command of the
harmonic signal. These techniques are based on frequency
domain, time domain and artificial intelligence. This paper
concerns in DQ0 extraction technique in time domain.
In this method, the load current is converted into the rotating
frame using (Park transformation) in the presence of zero
sequence current due to the current passing in the neutral as in
equation (3). To transfer from ABC frame to DQ0 rotating
frame, phase lock loop (PLL) is required to generate theta, sine
and cosine signals needed to be synchronized with the various
phase to neutral voltages as shown in Figure3.
With a simple high-pass filter, the DC part can be easily
removed from

i Lde

and

i Lqe

and the remaining can be

transformed into its previous frequency with a reverse

transformation. The oscillating parts of

i Lde and

i Lqe

represents the harmonic content of the load current as in equation

(4) .
IJER@2016

The two oscillating parts of the load current in the rotating frame
represents the reference current of the active filter. Then it is
transformed to the ABC frame as in equation (5).

Figure3. Theta generation using PLL.

III. DC-link voltage control
Voltage control require feedback control loop that is responsible
to keep its voltage stable at the reference value with minimum
error. The actual DC voltage VDCact is compared with a
reference VDCref (the reference voltage the capacitor must be
charged to). The error signal obtained from this comparison is
used to control the voltage. The amplitude of the current that is
required from the system to charge the capacitor when its voltage
decreases than the reference voltage (

I dc

) is evaluated

using equation (6):

I dc =Gce=Gc (VDC ref VDC act . )

(6)

Where; Gc represents a PI or fuzzy gain. This current will form a

part of the reference current that the inverter must follow in case
of harmonic compensation
In this paper, two controllers are used in maintaining constant
DC link.
IV. PI controller
The term tuning is used to describe the process of selecting the
optimum controller setting in order to obtain the best
performance from PI controller. The most widely used methods
are Ziegler and Nichols. They assumed one procedure called
ultimate cycle method which is based on using results from a
closed- loop test. With a control system employing a PI
controller, KP, KI have to be selected. Such selection determines
the response of the control system to the inputs [11]. PI
controller will eliminate forced oscillations and steady state error
resulting in operation of on-off controller.
The difference between the actual and set reference DC
voltage,which is the error, is passed through a PI controlller
which converts this difference into the required charging current

I dc to the capacitor to compansate the inverter losses as

doi : 10.17950/ijer/v5s12/1216

Page 949

International Journal of Engineering Research

Volume No.5, Issue No.12, pp : 948-953

ISSN:2319-6890 (online),2347-5013(print)
1 December 2016

shown in In Figure4. the reference DC voltage is choosen 600V

to be larger than the highest Vline in the system.

I dc

is

considered a peak value of the supply current, which is

composed of two components:
1. Fundamental active power component of load current.
2. Loss component of SAPF, to maintain the average capacitor
voltage to a constant value.

membership. Five membership functions are for output and

characterized using triangular and trapezoidal membership
functions [13]-[15]. These memberships are shown in Fig 6(a)
and 6(b).

Table 1. Fuzzy controller Rule Base

I ref in de frame

Figure4. Idc control using PI controller.

V. Fuzzy logic controller
The advantages of fuzzy controllers over conventional
controllers are that they do not require an accurate mathematical
model, can work with imprecise inputs and more robust. The
Mamdani type is used. The fuzzy controller is based on two
inputs; error and its derivative and one output, which are the
command signal to fuzzy controller.
Figure5 shows the internal structure of the control
circuit. The actual capacitor voltage is compared with a set
reference value. The error signal is then processed through a
Fuzzy controller, which contributes to the zero steady error in
tracking the reference voltage signal.

Figure 6(a). Input Variable Error E and dE Gaussian M.F.

error

Figure 6(b). Output "Id.c" Normalized M.F.

VI. Simulation results of the proposed system
de/dt

I ref in de frame

Figure5. Idc control using Fuzzy Logic controller.

Design of control rules:
The fuzzy control rule design involves defining rules that relate
the input variables to the output model properties. As fuzzy logic
controller is independent of system model, the design is mainly
based on the intuitive feeling for, and experience of the process.
The rules are expressed in English like language with syntax
such as If {error E is X and change of error E is Y} then
{control output is Z}. For better control performance finer fuzzy
petitioned subspaces {big negative, negative, zero, positive, and
big positive} are used and summarized in Table 1. The two
inputs, error and change in error, are three memberships;
negative, zero and positive and characterized using Gaussian
IJER@2016

1. PI controller
The simulation tests were obtained by using the computational
tool MATLAB/SIMULINK to verify the proposed system as
shown in Figure7. The system parametes are shown in Table 2.
The simulated system is consists of an A.C source,SAPF and
unbalanced load. This load is consists of three single phase diode
recifier with differnent connected loads. The SAPF is connected
to the system through an inductor. The source currents of phases
A,B,C and neutarl currents before compansation are shown in
Figure8,9,10 and 11. It is clear that the unbalaned loads caused a
high current zero sequence current value passing through the
neutral wire.The three source currents are highly distorted.
Different connected loads to the phases causes unbalnance
between them as shown in Figure12

doi : 10.17950/ijer/v5s12/1216

Page 950

International Journal of Engineering Research

Volume No.5, Issue No.12, pp : 948-953

ISSN:2319-6890 (online),2347-5013(print)
1 December 2016
20
15

Current in Ampere

10
5
0
-5
-10
-15
-20
0.2

0.21

0.22

0.23

0.24

0.25
Time in Seconds

0.26

0.27

0.28

0.29

0.3

Figure10. Source current of phase C before compansation.

30

Current in Am peres

20

Table 2. The studied system parameters

Va

10

-10

-20

Vb

-30
0.03

0.04

0.05

0.06
Time in Seconds

Vc

0.07

0.08

0.09

Figure11. Neutral current before compansation.

25
20

VSI1
Hysterisis
PWM

IA
IB

Current in Ampere

15

IC

10
5
0
-5
-10

Figure7. Overall proposed simulated system.

-15
-20
0.04

20

0.045

0.05

0.055

15

0.065

0.07

0.075

0.08

Figure12. Unbalanced source currents.

10

C urren t in A m pere

0.06

Time in Seconds

5
0
-5
-10
-15
-20
0.2

0.21

0.22

0.23

0.24

0.25
Time in Seconds

0.26

0.27

0.28

0.29

0.3

Figure8. Source current of phase A before compansation.

20
15

Current in Ampere

10
5
0
-5
-10

The compansated currents of the shunt active power filter in each

phase are shown in Figure 13,14 and15. These current will be
injected to the system at PCC to reshape the source current to
pure sine wave. Notice that the compansating current of phase B
is sinusodal wave and this is excepected due to its resestive load.
This compansating current will be added to the source current of
phase B in order to maintain balance between all the source
currents. The performance of the hysteresis controller appears in
the above compensating current curves. The controller successes
in tracking the calculated reference filter current. The proposed
system makes an excellent mitiagtion in the current harmoinc in
the source current. The source and neutral currents after
compansation is shown in figures 16,17,18 and 19.
60

-15

0.21

0.22

0.23

0.24

0.25
Time in Seconds

0.26

0.27

0.28

0.29

40

0.3

Figure9. Source current of phase B before compansation.

C u rren t in A m p ere

-20
0.2

20

-20

-40

-60
0.03

IJER@2016

doi : 10.17950/ijer/v5s12/1216

0.04

0.05

0.06
Time in Seconds

0.07

0.08

Page 951

0.09

International Journal of Engineering Research

Volume No.5, Issue No.12, pp : 948-953

ISSN:2319-6890 (online),2347-5013(print)
1 December 2016

Figure13. Compansating current of phase A.

60

40
40

20

C u rren t in A m p ere

20

Curent in Am peres

30

10
0

-20

-10

-40

-20
-30

-60
0.03

-40

0.04

0.05

0.06
Time in Seconds

0.07

0.08

0.09

-50
-60
0.03

0.04

0.05

0.06
Time in Seconds

0.07

0.08

0.09

Figure18. Source current of phase C after compansation.

Figure14. Compansating current of phase B.

30

20

Current in Am peres

It is clear that the proposed SAPF succeeded in mainating

balance between different phases as shown in Figure 20. The
steady state current value is the same in phase A,B and C and
equals to 30 Ampere. The current in the neutral is decreased to a
very low value nearly equals to 4 ampere.

10

-10

-20
40

-30
0.03

0.04

0.05

30

0.06
Time in Second

0.07

0.08

0.09

Figure19. Neutral current after compansation.

10
0
-10

50

-20

40

-30

30

-40
0.03

0.04

0.05

0.06
Time in Seconds

0.07

0.08

0.09

Figure15. Compansating current of phase C.

CurrentinAmpere

C u rren t in A m p eres

20

IB

IA

IC

20
10
0
-10
-20

60

-30
-40

40

Current in Amperes

-50
0.04

0.05

0.055

0.06
Time in Seconds

0.065

0.07

0.075

0.08

Figure20. Balanced source currents .

-20

-40

-60
0.03

0.04

0.05

0.06
Time in Seconds

0.07

0.08

0.09

Figure16. Source current of phase A after compansation.

40
30
20

Current in Am pere

0.045

20

10
0

Table 3 summarized the THD of the source currents before and

after compensation. It can be seen in figure 21 that the PI
controller has an very good performance in maintaing the
reference value. It takes nearly three complete cycles to reach the
steady state reference value (600V). This reflects that the P
(proportional) value is better chosen. At 0.09 second the steady
state error is very small and this reflects that the I (integral) value
is perfect. The value of the KP=0.6, KI=3.5. The tuning of these
values is using Ziegler and Nichols [11].

-10
-20
-30
-40
-50
0.04

0.045

0.05

0.055

0.06

0.065

0.07

0.075

0.08

0.085

0.09

Time is Seconds

Figure17. Source current of phase B after compansation.

Table 3. The THD of phases using PI controller

IJER@2016

doi : 10.17950/ijer/v5s12/1216

Page 952

International Journal of Engineering Research

Volume No.5, Issue No.12, pp : 948-953
600

D C V o lta g e in V o lts

500

400

300

200

100

-100

0.01

0.02

0.03

0.04

0.05

0.06

Time in Seconds

0.07

0.08

0.09

0.1

Figure21. Voltage of the PI DC link.

2. Fuzzy logic controller

The simulation was repeated using fuzzy to study the effect of
changing the DC controller on maintaing the DC reference
value.The controller scucced in tracking the reference signal and
compensating current tracks the reference calculated one. We
foucued on the DC voltage curve shown in Figure22 and on the
THD values of the source currents before and after compansation
as shown in Table 4 .
700
600

VDCinVolts

500
400
300
200
100
0
-100

0.01

0.02

0.03

0.04

0.05

0.06

0.07

0.08

0.09

0.1

Time in Second

Figure22. Voltage of the fuzzy DC link.

Table 4. The THD of phases using fuzzy controller

It can be seen in Figure 22 that the Fuzzy controller has an
excellent performance in maintaing the reference value. It has
fast response as it takes only one cycle and half to reach the
steady state reference value (600V). After 0.05 second the
steady state error is nearly null.
VII. Conclusion
In this paper, four wire SAPF based on DQ0 strategie for
reference currents extraction is considered. The three-phase
voltage source inverter using Hysteresis PWM succeeded in
IJER@2016

ISSN:2319-6890 (online),2347-5013(print)
1 December 2016
tracking the three-phase reference currents waveforms. The
proposed
system
had
been
implemented
using
MATLAB/Simulink. PI and fuzzy controllers are used for
maintaing constant DC voltage under unbalanced load
conditions. SAPF using PI controller mitigate harmonics and
decreases THD in phases A,C to 5.84% and 7.358%.
Fuzzy controller is used and simulation results demonstrate that
even if the supply current is unbalanced , the performance of
fuzzy controller showed better compensation capabilities in
terms of THD than with PI. THD decreases to 5.118% in phase A
and 6.61% in phase B.
SHAF using fuzzy controller has been found to be closer to the
IEEE 5191992 standard recommendations on harmonic levels,
making it easily adaptable to more severe constraints. The
proposed SAPF using both controllers has compensated the
neutral harmonic currents and the DC bus voltage is almost
maintained at the reference value under all disturbances, which
confims the effectiveness of both controllers.
References:
Rudnick, H., Dixon, J. , Moran, L.: Delivering Clean and Pure Power.
In: IEEE Power and Energy Magazine, Vol. 1 (2003), Issue 5,
September/October 2003, p. 32-40.
Mead, A., Nassir, A., Bekiarov, S. B.: Uninterruptable Power Supply
and Active Filters, CRC PRESS, Washington, D.C, 2005
Dugan, R. C., Santoso, S., McGranaghan, M. F.: Electrical Power
System Quality , Second Edition, McGraw Hill, New York, 1996.
Singh, B., Al-Haddad, K.: A Review of Active Filters for Power Quality
Improvement. In: IEEE Transactions on Industrial Electronics, Vol. 46
(1999), No. 5, October 1999, p.960 - 971.
Ise, T., Hayashi, Y., Tsuji, K.: Definitions of Power Quality Levels and
the Simplest Approach for Unbundled Power Quality Services.In:
Proceedings of the 9th International. Conference on Harmonics and
Quality of Power, Oct. 1-4, 2000, Orlando, Vol. 2, p.385 - 390.
Jiang , J., Zhang, B.H., Hao, Z.G., Klimek, A.: Study on Factors
Affecting the Harmonics of large scale rectification. In:, Proceeding of
the 45th International Universities Power Engineering Conference
UPEC, Aug. 31-Sep.3, 2010, p.1-5.
IEEE Standard.519-1992, IEEE Recommended Practices and
Requirements for Harmonic Control in Electric Power Systems, 1992.
Das, J.C.: Passive filters - Potentialities and Limitations. In: IEEE
Transactions on Industry Applications, Vol.40 (2004), Issue: 1,
January/February 2004, p.232- 241.
Peng, F. Z., Akagi, H.: A New Approach to Harmonic Compensation in
Power Systems-A Combined System of Shunt Passive and Series active
Filters. In: IEEE Transactions on Industry Applications, Vol.26 (1990),
No.6,November December 1990, p.983-990.
Zobaa, A.F., Abdel Aziz, M. M., Abdel Aleem, S. H. E.: Comparison of
Shunt-Passive and SeriesPassive Filters for DC Drives Loads. In:
Electr. Pow. Compo. Sys. Vol.38(2010), No.3, March 2010, p. 275-291.
Mamdouh, M. M., Mansour, A. A., N.F.Nashed, M., F.Zobaa, A., Abou
Elzahab, E. E.: Hysteresis PWM of Shunt Active Power Filter Based on
Prediction Harmonic Current Technique. In: Journal of Next
Generation, JNIT, Vol.5 (2014), No.2, May 2014, p.20-31
Mamdouh, M. M., Mansour, A. A., N.F.Nashed, M., F.Zobaa, A., Abou
Elzahab, E. E.: Comparison between Hysteresis and Space Vector
PWM of Shunt Active Power Filter Based on Prediction Harmonic
Current Extraction. In: Journal of Next Generation, JNIT, Vol.5 (2014),
No.2, May 2014, p.9-19
Mikkili, S., Kumar, A.: Mitigation of Harmonics using Fuzzy Logic
Controlled Shunt Active Power Filter with Different Membership
Functions by Instantaneous Power Theory. Proceeding of the Students

doi : 10.17950/ijer/v5s12/1216

Page 953

International Journal of Engineering Research

Volume No.5, Issue No.12, pp : 948-953

xiv.

Conference on Engineering and Systems SCES, March 16-18, 2012,

India, p.1-6.
Helal ,A. A., Zakzouk, N. E., Desouky, Y. G.: Fuzzy Logic Controlled
Shunt Active Power Filter for Three-phase Four-wire Systems with
Balanced and Unbalanced Loads. In: World Academy of Science,
Engineering and Technology, Vol.58(2009), p.621-626

IJER@2016

ISSN:2319-6890 (online),2347-5013(print)
1 December 2016
Musa, A.S., Radzi, M.A., Hisham, M,H.: Fuzzy Logic Controller Based
Three Phase Shunt Power Filter for Harmonics Reduction. In:
Proceeding of the IEEE conference of Energy Conversion CENCON,
Oct.13-14, 2014, Malaysia, p.371 - 376

doi : 10.17950/ijer/v5s12/1216

Page 954