PI and fuzzy logic controllers for shunt active power filter

ISA Transactions 51 (2012) 163–169
Contents lists available at SciVerse ScienceDirect
ISA Transactions
journal homepage: www.elsevier.com/locate/isatrans
PI and fuzzy logic controllers for shunt active power filter — A report
Karuppanan P∗
, Kamala Kanta Mahapatra
Department of Electronics and Communication, National Institute of Technology, Rourkela-769 008, India
a r t i c l e i n f o
Article history:
Received 8 March 2011
Received in revised form
11 September 2011
Accepted 12 September 2011
Available online 5 October 2011
Keywords:
Active power filter
PI-controller
Fuzzy logic controller
Adaptive-fuzzy hysteresis current
controller
Harmonic currents
a b s t r a c t
This paper presents a shunt Active Power Filter (APF) for power quality improvements in terms of
harmonics and reactive power compensation in the distribution network. The compensation process is
based only on source current extraction that reduces the number of sensors as well as its complexity.
A Proportional Integral (PI) or Fuzzy Logic Controller (FLC) is used to extract the required reference
current from the distorted line-current, and this controls the DC-side capacitor voltage of the inverter. The
shunt APF is implemented with PWM-current controlled Voltage Source Inverter (VSI) and the switching
patterns are generated through a novel Adaptive-Fuzzy Hysteresis Current Controller (A-F-HCC). The
proposed adaptive-fuzzy-HCC is compared with fixed-HCC and adaptive-HCC techniques and the superior
features of this novel approach are established. The FLC based shunt APF system is validated through
extensive simulation for diode-rectifier/R–L loads.
© 2011 ISA. Published by Elsevier Ltd. All rights reserved.
1. Introduction
Recently a lot of research is being encouraged for power quality
and custom power problems in the distribution system due to
non-linear loads [1,2]. In practical application most of the loads
are non-linear, such as power converters, SMPS, arc furnaces, UPS
and ASDs [3]. These non-linear loads are introducing harmonic
distortion and reactive power problems [4]. The harmonics in
the system induce several undesirable issues, such as increased
heating losses in transformers, low power factor, torque pulsation
in motors, poor utilization of distribution plant and also affects
other loads connected at the same Point of Common Coupling
(PCC) [5,6]. In recent times Active Power Filters (APFs) or Active
Power-Line Conditioners (APLCs) are developed for compensating
the harmonics and reactive power simultaneously [7,8]. The APF
topology can be connected in series or shunt and combinations
of both (unified power quality conditioners) as well as hybrid
configurations [9,10]. The shunt active filter is most popular than
the series active filter, because most of the industrial applications
required the current harmonics compensation [11,12].
The controller is the most significant part of the active power fil-
ter and currently various control strategies are proposed by many
researchers [13–15]. There are two major parts of the controller,
∗ Corresponding author. Tel.: +91 9442704585; fax: +91 661 2462999.
E-mail addresses: karuppanan1982@gmail.com (K. P), kkm@nitrkl.ac.in
(K.K. Mahapatra).
one is reference current extraction from the distorted line-current
and another is the PWM-current controller to generate switching
patterns for inverter [16,17]. Many control strategies are proposed
in the literature to extract the harmonic components. However,
the conventional PI controller requires precise linear mathematical
model of the system, which is difficult to obtain under parameter
variations and non-linear load disturbances. Another drawback of
the system is that the proportional and integral gains are chosen
heuristically [18,19]. Recently, Fuzzy Logic Controllers (FLCs) have
been used in various power electronic applications and also in ac-
tive power filters [20,21]. The advantage of FLCs over the conven-
tional controllers is that it does not need an accurate mathematical
model. It can handle nonlinearity and is more robust than conven-
tional PI or PID controllers [22,23].
Similarly, various current control techniques are proposed
for APF inner current control loop, such as triangular current
controller, sinusoidal-PWM, periodical-sampling controller and
hysteresis current controller [24,25]. The Hysteresis Current
Controller (HCC) method attracts researchers’ attention due to
unconditional stability and simple implementation [26]. However,
this control scheme exhibits several unsatisfactory features such
as uneven switching frequency where the switching frequency
varies within a particular band limit [27]. Adaptive-HCC (A-HCC)
overcomes these fixed-HCC demerits. The adaptive-HCC changes
the bandwidth according to instantaneous compensation current
variation [28–31]. However, the adaptive-HCC is having more
switching losses due to high frequency switching. These problems
can be overcome by the proposed Adaptive-Fuzzy-HCC (A-F-HCC)
in this paper. We focus on an adaptive-fuzzy-hysteresis current
0019-0578/$ – see front matter © 2011 ISA. Published by Elsevier Ltd. All rights reserved.
doi:10.1016/j.isatra.2011.09.004

164 K. P, K.K. Mahapatra / ISA Transactions 51 (2012) 163–169
Fig. 1. Shunt active power filter system.
control scheme, where the hysteresis-bandwidth is calculated
with the help of fuzzy logic [32–34]. This controller maintains
the modulation frequency constant that facilitates reduction of
switching loss. Furthermore, uneven switching does not occur
that facilitates reduction in the high frequency ripples in the
compensating current as well as load current. Therefore this
control strategy for inner current loop improves the PWM–VSI
performance and makes APF substantially.
This paper presents a shunt APF that uses PI or fuzzy logic
controller for reference current extraction. The active filter is
implemented with VSI and switching patterns are generated
from the proposed novel adaptive-fuzzy-HCC. The shunt APF
system is validated through extensive simulation under non-
linear load conditions. Comparative assessments of the different
PWM-current controller are presented. In Section 2, the design of
shunt APF system architecture is presented. Section 3 describes
about control strategies for reference current extraction and PWM-
current controller proficiencies. Section 4, system performance has
been modeled and analyzed from the obtained results under non-
linear load conditions. Finally, Section 5 describes the conclusions
of this work.
2. Design of shunt APF system
Shunt APF is connected in the distribution-grid at PCC through
filter inductance that is shown in Fig. 1. The filter inductance
suppresses the harmonics caused by the switching operation of the
power inverter. The current harmonic compensation is achieved by
injecting equal but opposite current harmonic components at PCC,
there by canceling the original distortion and improving the power
quality on the connected power distribution system [8].
The instantaneous source current is represented as
is(t) = iL(t) − ic (t). (1)
The instantaneous source voltage is
vs(t) = Vm sin ωt. (2)
The nonlinear load current contains the fundamental component
and harmonic current components, which is represented as [20]
iL(t) =
∞−
n=1
In sin(nωt + Φn)
= I1 sin(ωt + Φ1) +

∞−
n=2
In sin(nωt + Φn)

. (3)
Fig. 2. Block diagram of the control strategy.
The instantaneous load power can be computed from the source
voltage and load current and the calculation is given as
pL(t) = is(t)∗
vs(t)
= Vm sin2
ωt∗
cos ϕ1 + VmI1 sin ωt∗
cos ωt∗
sin ϕ1
+ Vm sin ωt∗

∞−
n=2
In sin(nωt + Φn)

= pf (t) + pr (t) + ph(t). (4)
This load power contains fundamental (active power), reactive
power and harmonic power. From this Eq. (4), the real (fundamen-
tal) power drawn from the load is
pf (t) = VmI1 sin2
ωt∗
cos ϕ1 = vs(t)∗
is(t). (5)
If the active power filter provides the total reactive and harmonic
power, the source current is(t) will be in phase with the utility
voltage and sinusoidal. The three-phase source currents after
compensation can be expressed as
i∗
sa(t) = pf (t)/vs(t) = I1 cos ϕ1 sin ωt
= Imax sin ωt. (6)
Similarly,
i∗
sb = Imax sin(ωt − 120°) (7)
i∗
sc = Imax sin(ωt + 120°). (8)
This peak value of the reference current Imax is estimated by
regulating the DC-bus capacitor voltage of the inverter using PI or
fuzzy logic controller.
3. Control strategies
The block diagram of the proposed control system is shown in
Fig. 2 that consists of two parts. One is reference current extraction
controller using unit current vector along with PI or fuzzy logic
controller. Another is PWM-current controlled voltage source
inverter switching control method using fixed-HCC or adaptive-
HCC or adaptive-fuzzy-HCC.
3.1. Reference current extraction controller
The reference current is extracted from distorted line-current
using unit current vector with PI or FLC. Conventional PI and PID
controllers have been used to estimate the required reference
current, and to control the dc-bus capacitor voltage of the
inverter [18,19]. When load is non-linear or reactive, it is

K. P, K.K. Mahapatra / ISA Transactions 51 (2012) 163–169 165
expected that the supply will supply only the active fundamental
component of the load current and the compensator would supply
the harmonic/reactive component. The outer voltage loop will
try to maintain the capacitor voltage nearly constant which is
also a mandatory condition for the successful operation of the
APF. The system losses are provided by the source in steady
state. The compensator supplies the harmonic power, which
manifests itself only on the reactive component of power. In
the transient conditions the load changes are reflected in the dc
capacitor voltage as an increase (or decrease) as capacitor absorbs
(or delivers) the excess (or deficit) power. This conservation
of energy philosophy is used to obtain the reference current
for compensator in this method. The perturbations in the
capacitor voltage are related to the perturbations in the average
power drawn by the non-linear load. This property is utilized
which facilitates extracting compensator reference and maintains
capacitor voltage nearly constant. This control algorithm is based
on sensing source current only (no need to sense load currents and
compensation currents) that reduces sensors and complexity.
3.1.1. Unit current vector
The source currents are sensed and converted into the unit sine
current(s) while corresponding phase angles are maintained. The
unit current vectors templates are represented as [21]
ia = sin ωt,
ib = sin(ωt − 120°) and
ic = sin(ωt + 120°).
(9)
The amplitude of the sine current is unity in steady state and in
the transient condition it may increase or decrease according to
the loads. The unit currents should be in phase with the source
voltages. These unit currents are multiplied with peak-amplitude
of the estimated reference current (using PI or FLC) which is used
to generate the desired reference currents.
3.1.2. PI-controller
The DC-side capacitor voltage is sensed and compared with
a reference voltage. The comparison result of the voltage error
e = Vdc,ref − Vdc at the nth sampling instant is used as an input
for Proportional Integral (PI)-controller. The error signal passes
through Low Pass Filter (LPF) to filter the higher order components
and passes only the fundamental. It transfer function is defined
as [18],
H(s) = KP + KI /s. (10)
The proportional gain is derived using KP = 2ζωnvC [KP = 0.7]
that determines the dynamic response of the DC-side voltage. The
damping factor ζ =
√
2 / 2 and natural frequency ωnv should
be chosen as the supply fundamental frequency. Similarly, the
integral gain is derived using KI = Cω2
nv[KI = 23] that determines
settling time and eliminates steady state error in the DC-side
capacitor voltage. This controller estimates the magnitude of peak
reference current Imax and controls the dc-side capacitor. This
estimated magnitude of peak current multiplies with output of unit
current vector, which generates the required reference currents to
compensate the harmonic components.
3.1.3. Fuzzy logic controller
Fuzzy logic control is deduced from fuzzy set theory; it was
introduced by Zadeh in 1965 [22]. In fuzzy set theory develops the
transitions between membership and non-membership function.
Therefore, limitation or boundaries of fuzzy sets can be undefined
and ambiguous and useful for approximate systems design. In
order to implement the fuzzy logic control algorithm of an APF in
Fig. 3. Fuzzy logic controller.
Table 1
Rule base table.
e(n) ce(n)
NB NM NS ZE PS PM PB
NB NB NB NB NB NM NS ZE
NM NB NB NB NM NS ZE PS
NS NB NB NM NS ZE PS PM
ZE NB NM NS ZE PS PM PB
PS NM NS ZE PS PM PB PB
PM NS ZE PS PM PB PB PB
PB ZE PS PM PB PB PB PB
a closed loop, the dc-bus capacitor voltage is sensed and compared
with the desired reference value. This computed error signal (e =
VDC−ref − VDC ) is passed through a LPF. The error signal e(n) and
integration of error signal or change of error signal ce(n) are used
as inputs for fuzzy processing as shown in Fig. 3.
Fuzzy logic is characterized by (1) Seven fuzzy sets (NB, NM, NS,
ZE, PS, PM, PB) for each input and output variables. (2) Triangular
membership function is used for the simplicity. (3) Implication
using Mamdani-type min-operator. (4) Defuzzification using the
height method.
Fuzzification: Fuzzy logic uses linguistic variables instead of
numerical variables. In a control system, the error between
reference signal and output signal can be assigned as Negative
Big (NB), Negative Medium (NM), Negative Small (NS), Zero (ZE),
Positive Small (PS), Positive Medium (PM), Positive Big (PB).
The processes of fuzzification convert numerical variables (real
number) convert to linguistic variables (fuzzy number).
Rule elevator: Conventional controllers like PI and PID have control
gains which are combination of numerical values. In case of fuzzy
logic controller uses linguistic variables instead of the numerical.
The basic fuzzy logic operations required for evaluation of fuzzy set
rules are AND (∩), OR (∪) and NOT (−) for intersection, union and
complement functions respectively, it is define as
AND—Intersection µA∩B = min[µA(X), µB(x)]
OR—Union µA∪B = max[µA(X), µB(x)]
NOT—Complement: µA = 1 − µA(x).
Defuzzification: The rules of fuzzy logic generate required output
in linguistic variables, according to real time requirements. The
linguistic variables have to be transformed to crisp number.
This selection of strategy is compromise between accuracy and
computational intensity.
Data and rule base: The database stores the definition of the
triangular membership function, which is required by fuzzifier and
defuzzifier. The rule-base stores the linguistic control rules which
are used by rule evaluator (decision making logic). The 49-rules are
used in this proposed controller as given in Table 1.
The output of the fuzzy controller estimates the magnitude of
peak reference current Imax. This current Imax takes response of the

Fig. 4. Diagram of hysteresis current control.
active power demand for harmonics and reactive power compen-
sation. This estimated magnitude of peak-current multiplies with
output of unit current vector and generates the required reference
currents. These reference currents compared with actual currents
to generate the inverter switching patterns using PWM-current
controller.
3.2. PWM–VSI current controller
The effectiveness of active power filter depends on the design
and characteristics of the current controller. There is various PWM-
current control strategies proposed for APF applications. But in
terms of quick current controllability and easy implementation,
HCC has highest rating among other methods such as sinusoidal-
PWM and triangular current controller [24,25]. This section
describes the fixed-HCC, adaptive-HCC and adaptive-fuzzy-HCC
for active power filter.
3.2.1. Fixed-hysteresis current controller
The fixed-hysteresis current controller is utilized indepen-
dently for each phase and directly generates the switching patterns
for the PWM-voltage source inverter [26]. It imposes a bang–bang
instantaneous control method that draws the APF compensation
current to follow its reference signal within a certain band limits.
The actual current iactual(t) is compared with iref (t) and the result-
ing error is subjected to a hysteresis controller to determine the
gating signals of the inverter as shown in Fig. 4.
If the error current exceeds the upper limit of the hysteresis
band, the upper switch of the inverter arm is turned OFF and
the lower switch is turned ON. As a result, the current starts to
decay. If the error current crosses the lower limit of the band, the
lower switch is turned OFF and the upper switch is turned ON.
As a result, the current gets back into the hysteresis band. This
switching performance is defined as
S =

OFF if iactual(t) > iref (t) + H
ON if iactual(t) < iref (t) − H

. (11)
The advantages of fixed-HCC are simple design and unconditioned
stability. However, this control scheme exhibits several unsatis-
factory features such as uneven switching frequency, difficult to
design the passive filters system to suppress ripples in compensat-
ing currents. This unpredictable switching function affects the APF
efficiency and reliability. Adaptive-hysteresis current controller
overcomes the fixed-HCC demerits. This adaptive-HCC changes the
bandwidth according to instantaneous compensation current vari-
ation.
3.2.2. Adaptive-hysteresis current controller
Fig. 5 shows current and voltage waves of the PWM-voltage
source inverter for phase a. The current ia tends to cross the lower
hysteresis band at point 1, where the transistor Q 1 is switched
ON. The linearly rising current (ia +) then touches the upper band
at point 2, where the transistor Q 4 is switched ON. The linearly
falling current (ia −) then touches the lower band at point 3. The
Fig. 5. Adaptive hysteresis current controller.
following equations can be written in the switching intervals t1
and t2 [27,28]
di+
a
dt
=
1
L
(0.5Vdc − Vs) (12)
di−
a
dt
= −
1
L
(0.5Vdc + Vs) (13)
where L = phase inductance; (i+
a ) and (i−
a ) are the respective rising
and falling current segments. From the geometry of Fig. 5, we can
write
di+
a
dt
t1 −
di∗
a
dt
t1 = 2HB′
(14)
di−
a
dt
t2 −
di∗
a
dt
t2 = −2HB′
(15)
t1 + t2 = Tc = 1/fc (16)
where t1 and t2 are the respective switching intervals, and fc is the
modulation frequency. Adding (14) and (15) and substituting (16)
we can write
di+
a
dt
t1 −
di−
a
dt
t2 −
1
fc
di∗
a
dt
= 0. (17)
Subtracting (15) from (14), we get
di+
a
dt
t1 −
di−
a
dt
t2 − (t1 − t2)
di∗
a
dt
= 4HB′
. (18)
Substituting (17) in (18), we get
di+
a
dt
(t1 + t2) − (t1 − t2)
di∗
a
dt
= 4HB′
. (19)
Substituting (13) in (17) and simplifying,
(t1 − t2) =
di∗
a /dt
fc (di+
a /dt)
. (20)
Substituting (20) in (19), it gives,
HB′
=

0.125Vdc
fc L

1 −
4L2
V2
dc

Vs
L
+ m
2

. (21)
Here, m = di∗
a /dt is the slope of reference current signals. The
hysteresis band HB′
can be modulated at different points of funda-
mental frequency to control the switching pattern of the inverter.
The bandwidth HB′
is applied to the variable HCC and it created by
s-functions in Matlab to produce switching patterns. This switch-
ing frequency of the inverter depends on the dc-side capacitor

Table 2
System parameters.
Parameters Values
Line to line source voltage (Vm) 440 V
System frequency (f ) 50 Hz
Source impedance: Source resistor (RS ) 1
Source inductor (LS ) 0.1 mH
Non-linear load: Diode rectifier 6-diode
Load resistor (RL) 20
Load inductor (LL) 100 mH
Filter:Inductor (LF ) 1 mH
Resistor (RF ) 1
DC-side capacitance (CDC ) 2100 µF
Reference voltage (VDC, ref ) 400 V
Power converter MOSFETs/diodes
voltage and line inductance of the APF configuration. This adaptive-
HCC is having more switching power losses due to high frequency
and this problem is mitigated by the proposed adaptive-fuzzy-HCC.
The adaptive-fuzzy-HCC calculates the bandwidth effectively with
the help of FLC and reduces the switching power losses by optimiz-
ing switching frequency.
3.2.3. Adaptive-fuzzy hysteresis current controller
The proposed adaptive-fuzzy-HCC calculates the hysteresis
bandwidth very effectively with the help of fuzzy logic controller.
The adaptive-HCC based band-width is calculated and is given in
the Eq. (21). From this equation, the hysteresis bandwidth HB′
is derived from the modulation frequency fc , supply voltage Vs,
dc-side capacitor voltage Vdc , and slope of the reference current
signals di∗
a /dt and decoupling inductance L of the active power
filter. The adaptive hysteresis band-width can be modulated as a
function of di∗
a /dt, Vs and Vdc . This adaptive hysteresis bandwidth
HB′
as an error signal E(HB′
) and change of error signal CE(HB′
)
are used as inputs for fuzzy logic processing. The adaptive-fuzzy-
hysteresis bandwidth (HB) is the output of the fuzzy controller.
The fuzzy logic rules are stored as linguistic variables, which
are required by the rule-evaluator. The 49-rules are used in this
adaptive-fuzzy-HCC to calculate the required band-width (HB) and
the rules are given in Table 1.
The output of FLC of HB (hysteresis bandwidth) can be
modulated at different points of the fundamental frequency cycle
to control the switching pattern of the inverter. For symmetrical
operation of all three-phases is denoted as HBa, HBb and HBc
of same value, but having 120° phase difference. The adaptive-
fuzzy-HCC based hysteresis-bandwidth HB should maintain the
modulation frequency fc constant. This controller reduces the
switching power losses and improves the PWM–VSI performances
for APF substantially.
4. Result and analysis
The performance of the proposed control strategy is evaluated
through Matlab/Simpower tools. The three-phase APF system
comprises six-IGBTs with diodes, a dc-bus capacitor, RL-filter,
compensation controller and switching pattern generator. The
compensation control process is developed from unit current
vector along with PI or FLC for extract the reference currents. The
switching patterns are generated from fixed-HCC, adaptive-HCC
and adaptive-fuzzy-HCC. These three-different types of current
control techniques are simulated and investigated. The designs
of the shunt APF system parameters value are given in Appendix
(see in Table 2).
Case 1 PWM-current controller.
The effectiveness of an active power filter basically depends
on the design and characteristics of the current controller. The
switching losses are increased in fixed-HCC method and currents
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
1
0.9
0.8
0.7
0.6
0.5
0.4
0.3
0.2
0.1
0
0 0.002 0.004 0.006 0.008 0.01 0.014 0.016 0.018 0.020.012
0 0.002 0.004 0.006 0.008 0.01 0.014 0.016 0.018 0.020.012
0 0.002 0.004 0.006 0.008 0.01 0.014 0.016 0.018 0.020.012
Fig. 6. Switching patterns (a) HCC (b) Adaptive-HCC and (c) Adaptive-fuzzy-HCC.
generated due to this approach contain high frequency ripples. The
PWM-current controller performance can be improved by using
the adaptive control technique. A novel A-F-HCC, based on the
adaptive control concept is used in this paper where hysteresis
band implementation is done using fuzzy logic; this approach
facilitates to optimize the PWM performance. The A-F-HCC utilizes
a look-up table method using the instantaneous supply voltage and
mains current reference slope as input variables and the hysteresis
band as an output variable to maintain the modulation frequency
constant. Fig. 6 illustrates the switching patterns of fixed-HCC,
A-HCC and A-F-HCC methods. The ripples in the currents are
reduced even if the switching frequency is increased; the A-F-HCC
facilitates increase of switching frequency as per the system
requirements. This A-F-HCC reduces the arbitrary switching and
current ripples, which improves the APF performance compared
to fixed-HCC and A-HCC.
Case 2 PI-controller.
The ac-main grid feeds diode-rectifier load. Shunt APF is
connected in parallel at PCC that injects the anti-harmonics.
PI-controller is used to estimate the peak amplitude of the
reference current and fixed-HCC or A-HCC and A-F-HCC controllers
used to generate switching patterns of the inverter. These
three-different types of current controllers are simulated and
investigated separately, but simulation of the Fig. 7 is focused on
A-F-HCC. The rectifier R–L load current or source current before
compensation is shown in Fig. 7(a). The shunt active filter supplies
the compensating current that is shown in Fig. 7(b). Consequently
current harmonic compensation is achieved by injecting equal but
opposite current harmonic components at PCC by canceling the
original distortion. The simulation result of source current after
compensation is presented in Fig. 7(c) that indicates the current
is sinusoidal. These figures are presented for A-phase only; other
phase waveforms are phase shifted by 120°.
Case 3 Fuzzy logic controller.
FLC-controller is used to estimate the magnitude of peak
reference current by controlling dc-side capacitor voltage of the
inverter. The simulation result of the 3-phase source voltage is
shown in Fig. 8(a) that indicates the source voltage is balanced.
The six-pulse diode-rectifier load current or source current before
compensation is shown in Fig. 8(b). The peak reference current is
multiplied with unit current vector templates and generates the

40
80
60
60
40
40
20
0
20
30
30
20
20
10
10
0
0.02 0.03 0.04 0.05 0.06 0.07 0.08 0.09 0.1
60
60
40
40
40
20
0
20
80
a
b
c
Fig. 7. Simulation results (a) source current before compensation (b) compensation
current (c) source current after active filter compensation.
Table 3
VDC settling time under various techniques.
Current controller VDC -settling time
PI FLC
Fixed-HCC (s) 0.29 0.25
Adaptive-HCC (s) 0.26 0.23
Adaptive-fuzzy HCC (s) 0.25 0.23
Table 4
Real and reactive power measured under various methods.
Current controller Without APF With APF
PI FLC
Fixed-HCC P = 10.85 kW P = 10.93 kW
Q = 118 VAR Q = 108 VAR
Adaptive-HCC P = 10.17 kW P = 11.14 kW P = 11.42 kW
Q = 269 VAR Q = 067 VAR Q = 053 VAR
Adaptive-Fuzzy HCC P = 11.27 kW P = 11.81 kW
Q = 059 VAR Q = 048 VAR
required reference-currents as shown in Fig. 8(c). That reference
currents compared with actual current and generates switching
patterns using fixed-HCC or A-HCC or A-F-HCC current controllers.
The controllers are tested individually; however, the simulation of
the Fig. 8 is focused on A-F-HCC. The shunt active filter produces
required compensation currents as shown in Fig. 8(d). The source
current after compensation is presented in Fig. 8(e) that indicates
the current is sinusoidal.
The APF is suppressing reactive/harmonic power and simulta-
neously improves power factor as shown in Fig. 8(f); the phase-a
voltage is in phase with current. The DC-side capacitance voltage
is controlled by PI or FLC. This controller reduces the ripples in the
capacitor voltage to certain level and makes settling time to a low
value (t = 0.23 s) and it is plotted in Fig. 8(g). This figure is for
FLC in the outer loop with A-F-HCC in the inner current loop. The
performance of the PI or FLC with different PWM-current control
based dc-side capacitor voltage settling time is measured and given
in Table 3.
The Real (P) and Reactive (Q ) power are calculated for different
cases and are given in Table 4. These results are measured under
diode-rectifier load using PI and FLC with three-different PWM-
current controller. These results indicate that reactive power is
suppressed by APF and hence improves power quality in the
distribution system.
The Fast Fourier Transform (FFT) is used to measure the order
of harmonics in the source current. The Total Harmonic Distortion
e
d
a
b
c
f
g
Fig. 8. Simulation results (a) source-voltage, (b) load current, (c) reference-current,
(d) compensation current (e) source current after compensation and (f) unity power
factor, (g) DC-side capacitor-voltages.
(THD) of the source current is measured without and with active
filter using PI and FLC along with different current controllers. The
THD measurement is compared for PI and FLC controller that is
presented in Table 5. The order of the harmonics versus magnitude
of the source current is plotted that is shown in Fig. 9(a). It is
shown that the supply current is distorted due to the presence
of fifth, seventh, eleventh and higher order of harmonic spectral
components and the measurement of THD is 26.67%. The Fig. 9(b)
is plotted using A-F-HCC current controller based active power
filter and it ensures that THD is less than 5% (THD = 1.72%) in
compliance with the harmonic standards.
The PI or FLC based active filter system is simulated and
validated under diode-rectifier load conditions. The adaptive-
fuzzy-HCC reduces the switching power losses and improves the
active filter performance in comparison with the fixed-HCC and

a
b
Fig. 9. Order of Harmonics (a) without APF and (b) with APF.
Table 5
THD is measured under various methods.
Current controller Without APF (%) With APF
PI (%) FLC (%)
Fixed-HCC 3.57 3.21
Adaptive-HCC 26.67 2.91 2.66
Adaptive-fuzzy HCC 2.38 1.72
the adaptive-HCC. Similarly FLC is giving better performance than
conventional PI-controller in terms of Vdc -settling time, reactive
power compensation and THD. The FFT analysis of the active filter
confirms that the THD of the source current is less than 5% that is
in compliance with IEEE-519 and IEC 61000-3 harmonic standards.
5. Conclusions
The fuzzy logic controller based three-phase shunt active power
filter facilitates current harmonics and reactive power compen-
sation in the distribution system. The compensation process uses
a PI or a fuzzy logic controller to extract reference-current while
controlling DC-side capacitor voltage of the inverter. The shunt
APF is implemented with PWM current controlled voltage source
inverter. This inverter switching patterns are generated from a
novel adaptive-fuzzy-hysteresis current controller. The adaptive-
fuzzy-HCC calculation of the hysteresis bandwidth is done with
the help of fuzzy logic. The shunt APF system is investigated un-
der fixed-HCC, adaptive-HCC and adaptive-fuzzy-HCC current con-
troller techniques. It is observed that adaptive-fuzzy-HCC reduces
the switching power losses compared to fixed and adaptive-HCC.
The measured total harmonic distortion of the source currents is
in compliance with IEEE 519 and IEC 61000-3 harmonic standards.
This proposed controller based APF system can be implemented in
field programmable gate array (FPGA) that would be attempted as
a future work.
Appendix
See Table 2.
References
[1] Gyugyi L, trycula ECS. Active AC power filters. In: Proc of IEEE/IAS annual
meeting 1976. 19-C: p. 529–35.
[2] Hosseini Mehdi, Ali Shayanfar Heidar, Fotuhi-Firuzabad Mahmoud. Reliability
improvement of distribution systems using SSVR. ISA Transactions 2009;48:
98–106.
[3] Duffey Christopher K, Stratford Ray P. Update of harmonic standard IEEE-
519: IEEE recommended practices and requirements for harmonic control in
electric power systems. IEEE Transactions on Industry Application 1989;25(6):
1025–34.
[4] Subjak Jr Joseph S, Mcquilkin John S. Harmonics — causes, effects, measure-
ments, and analysis: an update. IEEE Transactions on Industry Application
1990;26(6):1034–42.
[5] Emanuel Alexander E, Orr John A, Cyganski David, Gulachenski Edward M.
A survey of harmonic voltages and currents at the customer’s bus. IEEE
Transactions on Power Delivery 1993;8(1):411–21.
[6] Grady WM, Samotyj MJ, Noyola AH. Survey of active power line conditioning
methodologies. IEEE Transactions on Power Delivery 1990;5(3):1536–42.
[7] Singh Bhim, Al-Haddad Kamal, Chandra Ambrish. A review of active filter
for power quality improvements. IEEE Transactions on Industrial Electronics
1999;46(5):960–70.
[8] Akagi Hirofumi, hirokazu watanabe Edson, Aredes Mauricio. Instantaneius
power theory and applications to power conditioning. IEEE-press; 2007.
[chapter 3–4].
[9] Kale Murat, zdemir Engin O. Harmonic and reactive power compensation
with shunt active power filter under non-ideal mains voltage. Electric Power
Systems Research 2005;74:363–70.
[10] Habrouk MEl, Darwish MK, Mehta P. Active power filters: a review. IEE
Proceedings Electrical Power Applied 2000;147(5):403–13.
[11] Kumar Jain Shailendra, Agarwal Pramod, Gupta HO. A control algorithm for
compensation of customer-generated harmonics and reactive power. IEEE
Trans Power Delivery 2004;19(1):357–66.
[12] Rahmat Allah Hooshmand, Mahdi Torabian Esfahani. A new combined method
in active filter design in power systems. ISA Transactions 2011;50:150–8.
[13] Herrera RS, Salmeron P, Kim H. Instantaneous reactive power theory applied
to active power filter compensation: different approaches, assessment, and
experimental results. IEEE Transactions on Indstrial Electronics 2008;55(1):
184–96.
[14] Rahmat Allah Hooshmand, Mahdi Torabian Esfahani. Adaptive filter design
based on the LMS algorithm for delay elimination in TCR/FC compensators.
ISA Transactions 2011;50:142–9.
[15] Czarnecki Leszek S. Instantaneous reactive power p–q theory and power
properties of three-phase systems. IEEE Transactions on Power 2006;21(1):
362–7.
[16] Shatshat REl, Kazerani M, Salama MMA. Power quality improvement in 3-
phase 3-wire distribution systems using modular active power filter. Electric
Power Systems Research 2001;61:185–94.
[17] Massoud AM, Finney, Williams BW. Review of harmonic current extraction
techniques for an active power filter. In: International conference on
harmonies and quality of power. 2004. p. 145–59.
[18] Chaoui Abdelmadjid, Gaubert Jean Paul, Krim Fateh, Champenois Gerard.
PI controlled three-phase shunt active power filter for power quality
improvement. Electric Power Components and Systems 2007;35:1331–44.
[19] Karuppanan Pitchaivijaya, Kamala Kanta Mahapatra. A novel PLL with fuzzy
logic controller based shunt active power line conditioners. Journal of
Electrical and Electronics Engineering 2010;3(2):153–8.
[20] Jain SK, Agrawal P, Gupta HO. Fuzzy logic controlled shunt active power filter
for power quality improvement. IEE Proceedings Electrical Power Applied
2002;149(5):317–28.
[21] Saad S, Zellouma L. Fuzzy logic controller for three-level shunt active
filter compensating harmonics and reactive power. Electric Power Systems
Research 2009;79:1337–41.
[22] Raviraj VSC, Sen PC. Comparative study of proportional–integral, sliding mode,
and fuzzy logic controllers for power converters. IEEE Transactions on Industry
application 1997;33(2):518–24.
[23] Singh GK, Singh AK, Mitra R. A simple fuzzy logic based robust active power
filter for harmonics minimization under random load variation. Electric Power
Systems Research 2007;77:1101–11.
[24] Nabae Akira, Ogasawara Satoshi, Akagi Hirofumi. A novel control scheme for
current-controlled PWM inverters. IEEE Transactions on Industry Application
1986;1A-22(4):562–70.
[25] Brod DM, Novotny DM. Current control of VSI-PWM inverter. IEEE Transac-
tions on Industry Application 1985;21:562–70.
[26] Zeng Jiang, Yu Chang, Qi Qingru, Yan Zheng, Ni Yixin, Zhang BL, Chen Shousun,
Wu Felix F. A novel hysteresis current control for active power filter with
constant frequency. Electric Power Systems Research 2004;68:75–82.
[27] Kale Murat, Ozdemir Engin. An adaptive hysteresis band current controller for
shunt active power filter. Electric Power Systems Research 2005;73:113–9.
[28] Bose Bimal K. An adaptive hysteresis-band current control technique of a
voltage-fed PWM inverter for machine drive system. IEEE Transactions on
Industrial Electronics 1990;31(5):402–8.
[29] Ezoji E, Fazlali M, Ghatresamani A, Nopour M. A novel adaptive hysteresis band
voltage controller for dynamic voltage restorer. European Journal of Scientific
Research 2009;37(2):240–53.
[30] Dalvand H, Moghani JS, Talebi N. An adaptive hysteresis band current
controller for hybrid power filter. In: International conference on industrial
and information systems. 2006. p. 613–8.
[31] Belhaouchet N, Rahmani L. Development of adaptive hysteresis-band current
control of PWM three-phase AC chopper with constant switching frequency.
Electric Power Components and Systems 2009;33(10):583–98.
[32] Kadjoudj M, Abdessemed R, Golea N, Benbouzid ME. Adaptive fuzzy
logic control for high performance PM synchronous drives. Electric Power
Components and Systems 2001;29:789–807.
[33] Cecati Carlo, Corradi Stefano, Rotondale Nicola. A fuzzy logic-based adaptive
PWM technique. conference on IECON 1997; 3: p. 1142 – 47.
[34] Nounou Hazem N, Rehman Habib-ur. Application of adaptive fuzzy control to
ac machines. Applied Soft Computing 2007;7:899–907.

PI and fuzzy logic controllers for shunt active power filter

More Related Content

What's hot (20)

Viewers also liked (19)

Similar to PI and fuzzy logic controllers for shunt active power filter (20)

More from ISA Interchange (20)

Recently uploaded (20)

PI and fuzzy logic controllers for shunt active power filter