Page 1
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
247
HARMONIC MITIGATION FOR NON-LINEAR LOADS USING
THREE-PHASE FOUR WIRE UPQC CONTROL STRATEGY
Mr. Laith O. Maheemed1, Prof. D.S. Bankar
2
1 Bharatividyapeeth university .M.TECH Student, Electrical Department ,
COE, Pune, India
Email: [email protected] 2 Bharatividyapeeth university, Associated professor ,Electrical Department,
COE, Pune, India
Email: [email protected]
ABSTRACT
This paper presents three-phase four-wire active filter for power line conditioning (PLC) to
improve power quality in the DFIG wind turbine grid network. In addition to the power-factor
correction, load balancing and mitigation of voltage and current harmonics, it can regulate the
load voltage against voltage sag/swell and voltage dip in a three-phase four-wire distribution
system for different non-linear loads. The active power filter (APF) is implemented with PWM
based current controlled voltage source inverter (VSI). This VSI switching signals are generated
through proposed two-level hysteresis current controller (HCC) that achieves significant
reduction in the magnitude and variation of the switching frequency; The synchronous reference
frame (SRF) theory is used to get the reference signals for series and shunt active power filters
(APFs). The reference signals for the shunt and series APF of UPQC are derived from the control
algorithm and sensed signals are used in a hysteresis controller to generate switching signals for
shunt and series APFs. The UPQC is realized using two voltage source inverters (VSI) connected
back to back, to a common dc link capacitor. MATLAB/Simulink based simulations are obtained,
which support the functionality of the UPQC.
Keywords: DFIG, Active Power Line Conditioners (APLC), PI-Controller, Hysteresis Current
Controller (HCC), Harmonics, Power quality.
I . INTRODUCTION
Wind energy is the fastest growing and most widely utilized emerging renewable energy
technologies in electrical energy conversion systems at present [1]. This high penetration of wind
energy in the power system has been closely related to the advancement of the wind turbine
technology and the way of how to control. There are basically three types of generators that are
commonly used with commercial wind turbines. They are (1) fixed-speed system with squirrel-
INTERNATIONAL JOURNAL OF ELECTRICAL ENGINEERING &
TECHNOLOGY (IJEET)
ISSN 0976 – 6545(Print)
ISSN 0976 – 6553(Online)
Volume 3, Issue 1, January- June (2012), pp. 247-260
© IAEME: www.iaeme.com/ijeet.html Journal Impact Factor (2011): 0.9230 (Calculated by GISI)
www.jifactor.com
IJEET
© I A E M E
Page 2
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
248
cage induction generator, (2) variable-speed system with Doubly-Fed Induction Generator
(DFIG) (3) variable-speed system with a direct-drive synchronous generator. DFIG based
variable speed wind energy conversion systems are currently the most admired one, due to its
important advantages such as, high capacity with high energy efficiency, four-quadrant active and
reactive power controls and the small converter size with a rating of only 20%–30% of the rated
Wind turbine power [2]. The DFIG consists of a Wounded Rotor Induction Generator with the
stator windings directly connected to the constant frequency three-phase grid and with the rotor
winding connected to a bidirectional back-to-back PWM voltage Source Converter.
The ever-growing proliferation of power-switching devices for source conditioning and motion
control in single-phase and three-phase modern industrial applications has increased the
occurrence of unacceptable current harmonics levels in three-phase distribution systems. The
harmful and costly effects of harmonics have been discussed extensively in literature [3-6]. A
major effect of harmonic voltages and currents in rotating machinery (DFIG included) is
increased heating due to iron and copper losses at the harmonic frequencies. The harmonic
components thus affect the machine efficiency [7]. For instance, the fifth and seventh harmonics
can combine to produce a torsional stimulus on a generator rotor at the sixth harmonic frequency.
If the frequency of a mechanical resonance exists close to the frequency of electrical stimulus,
high-stress mechanical forces can be developed. Another generally greater concern is the flow of
harmonic currents in the rotor. The flow of each current in the stator will produce a magneto-
motive force in the air gap that will induce current flow in the rotor
of the machine. Just as each characteristic harmonic can be defined as being a positive or negative
sequence, the rotation of that harmonic will be either forward or backward with respect to rotor
rotation. The fifth harmonic will rotate in a backward direction (negative sequence), so a
harmonic current will be induced in the rotor with a frequency corresponding to the net rotational
difference between the fundamental air gap frequency and the fifth, i.e., the fifth plus one, or the
sixth harmonic. Since the seventh harmonic will rotate in a forward direction (positive sequence),
a harmonic current will be induced in the rotor with a frequency corresponding to the net
rotational difference between the seventh and the fundamental air gap frequency, i.e., the seventh
minus one, or the sixth harmonic. Thus, from a rotor heating standpoint, the fifth and the seventh
harmonics in the stator combine to produce a sixth harmonic current in the rotor. The 11th and the
13th harmonics act in the same manner to produce the 12th harmonic current in the rotor, and so
on with higher order harmonic pairs. There are two major concerns with these rotor harmonics:
(1) Resultant rotor heating; (2) Pulsating or reduced torques.
To solve these problems, passive power filters have been widely used for a long time. Passive
power filters consists of a combination of inductors and capacitors tuned to a certain frequency.
Although they are simple in structure and have a relatively low investment cost, they can cause
unwanted resonance and amplify harmonic currents. To overcome the disadvantage of passive
power filters and restrictions on their performance, research in active power filters has been
carried out actively. Active power filters can be classified as series or parallel by their system
configuration.
The combination of series and parallel active power filters is called the Unified Power Quality
Conditioner (UPQC). Although its main drawback is its high cost and complexity of control,
interest in UPQCs is growing due to its superior performance. UPQCs offer not only harmonics
elimination but also compensation for reactive power, load current unbalance, source voltage
sags, source voltage unbalance and power factor correction [2]. UPQC is mainly designed to
inject compensating current and voltage into the system, in order to mitigate the system
harmonics [3]. The UPQC have been studied and applied to regular three-phase power systems,
operated in 50/60 Hz [4, 5], however, there are no many applications of UPQC in single-phase
systems. This article is based on the steady state analysis of the behavior of UPQC with a
Page 3
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
249
distorted source voltage and a nonlinear load condition. Aim is to maintain the load bus voltage
sinusoidal and at desired constant level with a sinusoidal source current.
In this paper, a shunt active power filter is proposed to protect DFIG wind turbine from the
destroying effects of the current harmonics caused by the connection of nonlinear loads at the
point of common coupling (PCC). Simulation results using MATLAB/SIMULINK are shown to
validate the robustness and effectiveness of the SAPF to mitigate current harmonics.
II. DFIG MATHEMATIC MODEL
A doubly fed induction machine is a wound rotor with back-back converter in the rotor circuit. A
DFIG works as a generator or as a motor at both above and below the synchronous speed by
controlling the power injected into the rotor.
Fig 1 Detailed Configuration of 3P4WUPQC
In DFIG the rotor is supplied by PWM inverter, while the stator is directly connected to grid. The
rotor current exciting frequency is controlled as the wind velocity is changed. The frequency of
output power is fixed at grid frequency, which is given as follows:
ωs=p Ωm± ωr (1)
Where ωs is the grid electrical angular speed, m Ω is the mechanical angular rotor speed, r ω is
the electrical angular speed of rotor variables, and p is the number of pole pairs. In sub-
synchronous operation mode the sign in (1) is positive; otherwise it is negative in super-
synchronous operation mode.
Equation (1) establishes is the basis for VSCF.
The mathematical equations of the DFIG in terms of stator, rotor voltages and flux are given as
follows [3]:
Vsd=Rsisd –ωs φsq + φsd (2)
Vsq=Rsisq –ωs φsd + φsq (3)
Vrd=Rrird –(ωs-ω) φrq + φrd (4)
Vrq=Rrirq –(ωs-ω) φrd + φrq (5)
The direct and quadrature stator and rotor flux components are given as follows [4]:
φsd=Lsisd + Lmird (6)
φsq= Lsisq + Lmirq (7)
φsd=Lrird + Lmisd (8)
φrq= Lrirq + Lmisq (9)
The d-q steady-state equivalent circuit of the DFIG is depicted in Fig. 2 [5].
Page 4
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
250
DFIG Wind Turbine
The DFIG wind turbine adopted here is shown in Fig. 2. It consists of a DFIG driven by a wind
turbine and controlled on the rotor side through the Back-to-back PWM power converters. Back-
to-back PWM converters consist of two converters, the stator-side converter and rotor-side
converter, which are controlled independently of each other. The main idea is that the rotor-side
converter controls the active and reactive power by controlling the rotor current components,
while the stator-side converter controls the DC-link voltage and ensures a converter operation at
unity power factor (zero reactive power). Depending on the operating condition of the rotor, the
power is fed into or out of the rotor. In an over synchronous condition, power flows from the
rotor via the converter to the grid, whereas power flows in the opposite direction in a sub-
synchronous condition. In both cases, the stator feeds power into the grid at the point of common
coupling (PCC) through a transformer [8].
Fig. 1 shows the proposed studied system configuration. This system consists of a DFIG wind
turbine, grid supply, SAPF and nonlinear load all connected at the point of common coupling
(PCC). The shunt active power filter is composed of three parts. Three legs voltage source
converter (VSC) connected to the PCC through interfacing inductors, a DC link represented by a
capacitor and a control system. The nonlinear load is a three phase diode rectifier feeding RL
load.
4. Shunt Active Power Filter (SAPF) Shunt active power filter is a power converter utilized in order to compensate current disturbances
(harmonics, reactive power and unbalance). In order to meet quality enhancement constraints
proper control of its power switches is needed. Several topologies and configuration have been
introduced in the literature and in commercial implementations for this filter that highlight
different aspects of its compensation tasks. The most common topology of the shunt active power
filter is shown in fig. 1. Its main components are voltage source converter, DC bus (in our
situation is a capacitor), output passive filter and a control system. The most important objective
of the SAPF is to compensate the current harmonics generated by non linear loads. The reference
currents consists of the harmonic components of the load currents which the active filter must
supply [9]. These reference currents are fed through a controller to generate switching signals for
the power switching devices of the voltage source converter (VSC). Finally, the AC supply will
only need to provide the fundamental component for the non linear load.
Page 5
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
251
III. ROTOR SIDE CONVERTER CONTROL ALGORETHIM- HYSTERESIS
CURRENT CONTROL
In this paper, hysteresis control technique is used to control the current harmonics injected by
SAPF into the grid. According to reference current and the injected current, the hysteresis control
determines switching signals for the inverter gates. Hysteresis current control is based on error
signal between the injected current (If) and the reference current (Iref generated by p-q theory)
which produces proper control signals.
Assuming the DFIG is connected to the state power grid in which the grid voltage and frequency
is constant. Fixing the d-axis of the synchronous frame on the stator voltage vector, a stator
voltage oriented (SVO) control is obtained.
Thus, the vector of the stator voltage is:
Vs=Vsd +j0 (10)
According to (10), the active and reactive power output from the stator side of the DFIG can be
represented as:
Ps=Vsd isd (11)
Qs= -Vsd isq (12)
Substituting (2) in (11) and (3) in (12) respectively, the active and reactive powers can be derived
as follows:
Ps= (13)
Qs= (14)
As seen from (13) and (14), the active and reactive powers are related to rotor currents ird and irq
respectively. Therefore, the active and reactive power can be controlled via ird and irq respectively,
which is possible through the control of Vrd and Vrq.
There is Hysteresis Band (HB) above and under the reference current and when the difference
between the reference and inverter current reaches the upper (or lower) limit; the current is forced
to decrease (or increase) as shown in Fig.3.
Fig 3 Hysteresis control loop with duty cycle waveform
T1+T2=T= (15)
The proposed control algorithm assumes converting the two reference signals ird_ref and irq_ref
using Park’s transformation inverse into abc reference frame, then by comparing the three rotor
current signals ira_ref , irb_ref and irc_ref with actual rotor currents. The error signals issued from
comparison are applied to hysteresis controllers. The logical outputs of these controllers are the
switching signals of power transistor in RSC.
The proposed algorithm based on hysteresis controller is shown in Fig. 4.
Page 6
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
252
Where fc is the switching frequency and has an inversely proportional relation to HB [11]. In
comparison with other PWM methods, the hysteresis current control has a very fast response, a
simple operation and a variable switching frequency [12].
7. Simulation Results and Discussion
The proposed system configuration of Fig.1 has been simulated by Simulink of Matlab as it is
shown in Fig. 4. The line voltage at PCC is 380 V with line frequency of 50 Hz. The nonlinear
load is a three-phase diode rectifier with rating of 80 kVA. SAPF is connected to the system
through a three-phase link inductor with Lf = 0.1 mH and the dc bus capacitance is C=4.4 mF
with reference dc voltage of VCd=850 V. The VSC is a voltage source full-bridge IGBT based
inverter driven by hysteresis control. DFIG wind turbine of 500 kVA is connected at PCC.
The following waveforms show the high efficiency of SAPF for mitigation of harmonics. Fig. 5
and Fig. 6 show one phase Voltage waveform at PCC and its spectral decomposition before and
after harmonic compensation. From Fig. 5 we can see clearly that the PCC voltage is distorted
and the total harmonic distortion (THD) parameter is 7.86 % which is according to the IEEE 519-
1992 standard is not tolerable because it exceeds the limit of 5%. Voltage distortion in this case is
due to the passage of current harmonics through the impedance of the grid (Zs), that is why in its
spectral decomposition we find the same harmonics rank (5, 7, 11, 13, ….) as that found in the
current driven from PCC. So, this distortion will disappear when current harmonics are
compensated by SAPF as it is shown in Fig. 6. After the mitigation of the current harmonics, the
THD parameter of the PCC voltage is reduced from 7.86 % to 0.09 % and the spectral
decomposition shows a strong attenuation of the magnitude (MAG) of all harmonics rank.
Phase Current waveform driven from PCC and its spectral decomposition before and after the
connection of SAPF is illustrated by Fig.7 and Fig. 8 respectively. Fig. 7 shows a distorted
current waveform with high value of THD parameter reaching 26.87 % which is according to the
IEEE 519-1992 standard is not tolerable because it exceeds the limit of 5%. The corresponding
spectral decomposition shows a very important magnitude of all harmonics rank especially the
5th and the 7th.
Once SAPF is connected it injects the same current harmonics into the grid but with opposite
phase (Fig. 9) and consequently the current driven from PCC becomes practically sinusoidal as it
is shown in Fig. 8. After the compensation of the current harmonics, the THD parameter of the
current driven from PCC is reduced from 28.87 % to 0.91 % and the spectral decomposition
shows a strong attenuation of the magnitude (MAG) of all harmonics rank.
For the SAPF to work properly, it must have a good regulation of its DC-link voltage. The
purpose of the DC-link voltage controller is to preserve the DC-link voltage Vcd at its reference
Page 7
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
253
value Vcd ref = 850 V. This is accomplished by balancing the active power flow in the capacitor.
The performance of the DC-link voltage controller is verified by simulations in Fig. 10. The DC-
link voltage controller consists of a PI-controller, where the integral part reduces the steady state
error of the DC-link voltage. This implies a faster response to changes in the capacitor current
and thereby reduction of the DC-link voltage deviation during transients
VI. Grid Side Converter (GSC) CONTROL ALGORITHM
The proposed algorithm of GSC adopts the SVO technique to regulate DC-Link voltage and
achieve a unity power factor. This strategy leads to getting the following active and reactive
powers:
Prec=Vd id (17)
Qrec=-Vd iq (18)
Thus, the current command of q-axis controls the reactive power and it is obvious that the current
command of q-axis must be zero iq_Ref =0 for unity power factor operation. Whereas a current
command of d-axis controls the active power, and consequently controls indirectly the DC-link
voltage. From the above mentioned analysis, the d-axis must have two loops; inner one, which
employed hysteresis controller to regulate the d-axis current; the outer loop; which uses
proportional-integral controller to control the DC-bus voltage. The output of the PI controller
generates id_Ref.
The reference d-axis current, which is formed by PI controller and q-axis current, which set to
zero are both transformed to abc reference frame using the Park’s transformation inverse. The
error signals issued from the comparison between current reference values and actual ones are
applied to PWM controllers. The logical outputs of these controllers are the switching signals of
power transistor in GSC to maintain the desired currents. The proposed scheme is shown in fig. 6.
In terms of the steady-state condition, Vdq= Vd+ j0 if the d-axis of the reference frame is aligned
along the PCC voltage position. Assuming Vdq1= Vd1+ j Vq1 and neglecting the grid filter
resistance, then, the current flowing between the PCC and the GSC according to (3) is
Idq= - (19)
in which Xf stands for the grid filter reactance.
Fig 5: proposed grid side controller (GSC)
The following issues are considered in the design of the conventional nested-loop control system.
1) To prevent the converter from getting into the nonlinear modulation mode, a saturation
mechanism is applied to the output voltage of the controller if the amplitude of the
reference voltage generated by the inner current-loop controller exceeds the converter
linear modulation limit. The general strategy is to set a limitation on but keeps
unchanged as shown by (11) [13], [14], where and are d the q and
Page 8
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
254
components of the modified controller output voltage and Vmax is the maximum allowable
voltage. It is found that any other saturation mechanisms could cause more system
oscillations and unbalances.
Vmax . cos(
= Vmax . cos( (20)
2) To prevent the GSC from exceeding the rated current, the -axis current reference is
adjusted if the amplitude of the reference current generated by the outer control loop
exceeds the rated current limit. The general approach is keeping the d–axis current
reference unchanged to maintain dc-link voltage control effectiveness while modifying
the q-axis current reference to satisfy the reactive power or ac bus voltage support
control demand as much as possible as shown by (12) [13], [14]
=sign ( . Sqrt((2-(
2 (21)
The overall control structure of the GSC is shown by Fig. 4, which consists of a -axis loop for dc-
link voltage control and a d-axis loop for reactive power or grid voltage support control. Signal
processing technology is applied to the measured dc-link voltage and d-and q-axis currents to
prevent the high order harmonics from entering the controllers. The current-loop controller may
integrate PI, fuzzy and adaptive control technologies to improve the dynamic performance of the
GSC. The PI part of the controllers operates on a direct target control principle. The fuzzy and
adaptive parts of the controllers adjust the PI parameters based on the error, between the
controlled variable and its target value, and the change in error [18]. The initial values of the PI
current-loop controllers are tuned according to the fundamental intelligent control principle, i.e.,
minimizing the rms error between the reference and measured values [15].
In addition, a nonlinear programming strategy as shown below is utilized to prevent the GSC
from going over the rated current and to avoid the converter getting into a nonlinear modulation
mode, where irated is the rated GSC phase rms current and is the reference reactive power
absorbed from the grid by the GSC. The basic principle of the nonlinear programming
formulation is that under GSC rated current and linear modulation limits, the system should
operate to achieve the dc-link voltage control goal while minimizing the difference between the
reference and actual reactive power as much as possible
Minimize: | - | (22)
Subject to: Vdc= (23)
/(2 ) (24)
The nonlinear programming strategy is implemented in the following way. If | | generated by
the outer dc-link voltage and reactive power control loops exceeds the rated current limit, and
are modified by (12). If | | generated by the inner current control loops exceeds the converter
linear modulation limit, d-axis and q-axis the axis voltages are recalculated by (14). As it can be
seen, the recalculation does not change the q-axis control voltage so that the q-axis control
loop is not affected. Hence, according to (7), the effectiveness of active power or dc-link voltage
control is maintained. But, the recalculation makes the d-axis control voltage does not
follow the control voltage generated by the d–axis current-loop controller. Thus, the
effectiveness of the reactive power or bus voltage support control, according to (7), would be
affected. Under such conditions, the reactive power control is actually decided by the constraint
of the converter linear modulation requirement but not the control rule
Page 9
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
255
=sign( . ) = (25)
V. SIMULATION
All simulation studies presented in this paper have been done with MATLAB/ Simulink with use
of the SimPowerSystems Toolbox. The generators are represented by a model of the electrical
circuit and the mechanical part is neglected due to the small speed deviation during the time
period considered. Therefore, the simulations were carried out with constant rotor speed. The
IGBT converters are modeled as ideal switches with anti-parallel diodes. Distributed parameter
models are used for lines and the transformer models consider saturation effects, but no
hysteresis. Circuit breaker models are ideal and open exactly at the first current zero crossing
after the open command.
Fig 6 Simulation design of the system
For the simulation scenarios a 15kVA wind farm is modeled by two equivalent wind turbines in
scenario A and by one equivalent wind turbine. The performance of the three-phase four-wire
shunt APLC system is evaluated through Matlab programs in order to program and test the
system under unbalanced non-linear load conditions. The system parameters values are; Line to
line source voltage is 440 V; System frequency (f) is 50 Hz; DC-link capacitor C=5000 µF ;
Reference dc voltage 600 V; Interface inductor is 5 mH and 1 Ω full bridge rectifier load 168 + j
16 Ω.
The conventional power circuit of the voltage source inverter based active power filter connected
at the point of common coupling shown in Fig 6. The voltage source inverter has six power
transistors with freewheeling diodes and two energy storages capacitor on DC-side that is
implemented as a four-wire active power filter.
The source draws non-sinusoidal or harmonic current due to the non-linear load. This nonlinear
load current contains the fundamental signals and harmonic current components. Fig 7 & 8 shows
the unbalanced non-linear load current or source current before compensation. It is indicate that
the grid and wind generator source having voltage and current harmonic components due to the
non-linear load, which is diode rectifier.
Page 10
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
256
Fig 7. Simulation wind turbine side voltage and current without controller
Fig 8. Simulation grid voltage and current without controller
The total harmonic distortion (THD) of wind generator FFT analysis voltage and current
waveforms were as shown in Fig 9 & 10 and grid side voltage and current waveforms were
shown in Fig 11 & 12.
Fig 9. Showing Wind generator voltage harmonics due to non-linear load without controller
Fig 10. Showing Wind generator current harmonics due to non-linear load without controller
Page 11
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
257
Fig 11. Showing grid voltage harmonics due to non-linear load without controller
Fig 12. Showing grid current harmonics due to non-linear load without controller
The 3-phase source voltages are converted to the 3-phase unit current(s) while corresponding
phase angles are maintained. The unit current is defined as
(26)
These unit currents multiplied with peak reference current for generating the reference currents.
The proposed PI-control scheme estimates the peak reference current of an APF system. The two
storage DC-side capacitor voltage is sensed and compared with a reference voltage. The error,
e= Vdc_ref -Vdc (27) at the sampling instant is used as input for PI-controller.
The following waveforms show the high efficiency of SAPF for mitigation of harmonics. Fig. 9
to 12, shows voltage and current waveforms at wind side generator and grid side source and its
spectral decomposition before harmonic compensation.
From Fig. 9 and 11, we can see clearly that the voltage is distorted and the total harmonic
distortion (THD) parameter is about 56% and 58% which is according to the IEEE 519-1992
standard is not tolerable because it exceeds the limit of 5%. Voltage distortion in this case is due
to the passage of current harmonics through the impedance of the grid (Zs), that is why in its
spectral decomposition we find the same harmonics rank (5, 7, 11, 13, ….) as that found in the
current driven from PCC. So, this distortion will disappear when current harmonics are
compensated by SAPF as it is shown in Fig. 6. After the mitigation of the current harmonics, the
THD parameter of the wind generator side voltage and current is reduced from 56.56 % to 1.50%
and grid side voltage spectrum THD has decreased from 57.84% to 1.92% without and with
compensation as shown in Fig . The spectral decomposition shows a strong attenuation of the
magnitude (MAG) of all harmonics rank.
Page 12
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
258
Fig 13 Simulation wind turbine side voltage and current with controller
Fig 14. Simulation grid side voltage and current without controller
Fig 15. Showing Wind generator voltage harmonics due to non-linear load without controller
Fig 16. Showing Wind generator current harmonics due to non-linear load without controller
Page 13
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
259
Fig 17 Showing grid voltage harmonics due to non-linear load without controller
Fig 9. Showing grid current harmonics due to non-linear load without controller
By using the proposed controller, the harmonics which were produced because of non-linear load
were minimized by the wind generator system control circuit. This technique is very efficient in
decreasing voltage harmonics to a very great extent.
VI. CONCLUSION
A new current decomposition technique, based on frequency domain and SRF theory, with
indirect current control and reduced number of current sensors for prioritized selective
compensation of different power qualities and their combinations has been investigated for the
shunt APF of three-phase three-wire UPQC. A control strategy based on SRF theory is applied
for the control of the series APF of UPQC. The observed performance of the UPQC has
demonstrated the ability of the proposed control technique to selectively compensate the customer
generated harmonics, the total source current harmonics, unbalanced loading, reactive power and
voltage harmonics, based on priority to respect the limited power capacity of VSIs employed for
the shunt and series APFs. In addition to this, by mitigation of customer generated harmonics
only, the responsibility of the utility and customers at the PCC is attributed. It is also observed
that the proposed control scheme has a fast response and is able to maintain the voltage and
current harmonics levels, thus conforming to IEEE-519 standards. Further, the applied control
scheme is able to self support the dc bus voltage of back to back connected VSIs of the UPQC.
The control scheme of shunt APF has the advantage of flexibility in the selection of the power
quality indices for which the reference may be computed. In addition to this the shunt APF
compensates the current based distortions even under distorted utility voltages, hence the
operation of shunt and series APF are independent of each other. In case of a voltage sensitive
load, the series APF may be switched on to mitigate the voltage harmonics present in the load
voltages.
Page 14
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 3, Issue 1, January-June (2012), © IAEME
260
APPENDIX
The system parameters used are as follows:
Supply voltage and line impedance: 415V L-L, f=50 Hz,
Rs=0.1 Ω, Ls=0.05mH
Ripple Filter: R=7 Ω, C=5mF
DC bus capacitance: Cdc=3000mF
DC bus voltage of UPQC: Vdc=600V
Series Transformer: 250KVA, 1.1KV/5.5KV
Loads: 1) Three-Phase Rectifier Load with R=25Ω on dc side, and 2) Three single phase load
10KW, 6KVar (lagging) in each phase.
REFERENCES [1] A. Ghosh and G. Ledwich , Power Quality Enhancement Using Custom Power Devices,
Kulwer International Series in Engineering and Computer Science, 2002.
[2] N. G. Hingorani, “Introducing custom power,”in Proc. IEEE Spectrum, Vol. 32, pp. 41-48,
Jun.1995.
[3] A. Cetin, H.F. Bilgin, A. Acik, T. Demirci, K.N. Kose, A. Terciyanli, B. Gultekin, N. Aksoy,
B. Mutluer, I. C¸ adirci, M. Ermis, K. Ongan, and
N. Akinci , “Reactive power ompensation of coal conveyor belt drives by using D-STATCOMs,”
in Proc. IAS, pp.1731-1740, 2007.
[4] M. J. Newman, D. G. Holmes, J.G. Nielsen, and F. Blaabjerg, “A dynamic voltage restorer
(DVR) with selective harmonic compensation
at medium voltage level,” IEEE Trans. Ind. Appl., Vol. 41, pp.1744-1753, Nov./Dec. 2005.
[5] K. H. Kwan, P. L. So, and Y. C. Chu, “A harmonic selective unified power quality conditioner
using MVR with kalman filters,” in Proc. IPEC, pp.332-337, 2007.
[6] M. J. Newman and D. G. Holmes, “A universal custom power conditioner (UCPC) with
selective harmonic voltage compensation,” in Proc. IECON, Vol. 2, pp. 1261-1266, 2002.
[7] H. Akagi, Y. Kanazawa, and A. Nabae, “Generalised theory of the instantaneous reactive
power in three-phase circuits,” in Proc. IEEE and JIEE IPEC, pp. 821-827, 1983.
[8] Y. Komastu and T. Kawabata, “Experimental comparison of p-q and extended p-q methods
for active filter,” in Proc. EPE, Vol. 2, pp. 2.729-
2.734, 1997.
[9] M. Depenbork and V. Staut, “The FBD-method as tool for compensating total non-active
currents,” in Proc. IEEE Harmonics and Quality of Power, pp.320-324, 1998.
[10] L. S. Czarnecki, “Orthogonal decomposition of the currents in a 3-phase nonlinear
asymmetrical circuit with a non-sinusoidal voltage source,” IEEE Trans. Instrum. Meas., Vol. 37,
No. 1, pp. 30-34, Mar. 1998.
[11] F. P. Marafao, S. M. Deckmann, J. A. Pomilio, and R. Q. Machado, “Selective disturbance
compensation and comparisons of active filtering strategies,” in Proc. IEEE Harmonics and
Quality of Power, pp. 484-489,
2002.
[12] B. Singh, V. Verma, and J. Solanki, “Neural network-based selective compensation of
current quality problems in distribution systems,” IEEE Trans. Ind. Electron., Vol. 54, No. 1, pp.
53-60, Feb. 2007.