UNIFIED POWER QUALITY CONDITIONER FOR COMPENSATING POWER QUALITY PROBLEM AD by iaemedu

VIEWS: 0 PAGES: 19

									INTERNATIONAL JOURNAL OF ELECTRICAL ENGINEERING &
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME
                             TECHNOLOGY (IJEET)

ISSN 0976 – 6545(Print)
ISSN 0976 – 6553(Online)                                                          IJEET
Volume 4, Issue 4, July-August (2013), pp. 74-92
© IAEME: www.iaeme.com/ijeet.asp
Journal Impact Factor (2013): 5.5028 (Calculated by GISI)
                                                                               ©IAEME
www.jifactor.com




     UNIFIED POWER QUALITY CONDITIONER FOR COMPENSATING
        POWER QUALITY PROBLEM: ADAPTIVE NEURO-FUZZY
                 INTERFERENCE SYSTEM (ANFIS)

           Mr. M. Vishnu vardhan1*, Mr. N.M.G.Kumar1, Dr. P. Sangameswararaju2
                                1
                                    (EEE, SV University, Tirupati, India)
                                2
                                    (EEE, SV University, Tirupati, India)


ABSTRACT

        Nowadays, the power quality related issues are more concentrating in power electronic
devices. Hence, the electronic devices are affected by the PQ disturbances. Various power devices
are used for compensating the PQ disturbances. Unified power quality conditioner (UPQC) is one of
the power electronics devices that are used for enhancing the PQ. It consists of two voltage source
inverters (VSIs) and sharing with one DC link capacitor. The discharging time of DC link capacitor
is very high, and so it is the main problem in UPQC device. In this paper, an enhanced Adaptive
Neuro-fuzzy Inference System (ANFIS) based UPQC is proposed to eliminate this problem. Main
Objective of this paper is to improve the power quality problem compensating performance of
UPQC. The purpose of ANFIS is use to generate the discharging dc link voltage by bias voltage
generator. ANFIS permits the combination of numerical and linguistic data. The generated fuzzy
rules are then trained by using the neural network and we get a desired output from the interference
system. The output of ANFIS is injected to the line by the proposed UPQC system. Then, the power
quality problem compensating performance of proposed ANFIS based UPQC is analyzed. The
analyzed results are compared with Neuro-fuzzy controller (NFC), artificial neural network (ANN),
fuzzy logic controller (FLC) based UPQC system.

Keywords: ANFIS, UPQC, Neural Network, Fuzzy Logic controller and Power Quality
Disturbance.

1. INTRODUCTION

        PQ studies have emerged as a significant topic because of the extensive use of sensitive
electronic equipments [6]. A broad definition of power quality that includes the definitions of
technical quality and supply continuity states that the limits specified in the standards and regulations
should not be exceeded by electrical PQ or in other words frequency, number and interval of

                                                    74
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

interruption, interruption in voltage, sine waveform and voltage unbalance [13]. Nowadays power
quality is definitely a big issue and the inclusion of advanced devices, whose functioning is
extremely sensitive to the quality of power supply, makes it especially important [17]. Due to the
increasing anxiety over supplying pure electrical energy to the consumers in the availability of non-
sinusoidal waveforms, PQ has gained much interest in recent years. [9].
        Huge number of non-linear loads and generators on the grid, especially systems based on
power electronics like variable speed drives, power supplies for IT-equipment and high efficiency
lighting and inverters in systems producing electricity from renewable energy sources have made
electrical energy systems, voltages and specifically currents extremely irregular [8]. Degradation or
impairment can occur in the electrical equipments connected to the system as a result of poor PQ
[12].
        The increased anxiety has resulted in measurement of changes in PQ, analysis of power
disturbance characteristics and generation of solutions to the PQ problems [4]. Any apparent problem
in voltage, current that gives rise to any frequency variations leading to breakdown or malfunction of
customer equipment is termed as the PQ problem [1] [10]. A PQ problem can be caused by several
events. As a switching operation within the facility to a power system may be associated with the
cause of a fault event located hundreds of miles away from that place, analysis of these events is
frequently difficult [2]. PQ problems include short disruptions, long disruptions, voltage sags and
swells, harmonics, surges and transients, unbalance, flicker, earthing defects and electromagnetic
compatibility (EMC) problems [5]. Undesirable effects like extra heating, intensification of
harmonics because of the existence of power factor correction capacitor banks, decrease of
transmission system efficiency, overheating of distribution transformers, disoperation of electronic
equipment, improper functioning of circuit breakers and relays, incorrectness in measuring device,
interference with communication and control signals etc are caused by these PQ problems [7] [11].
        Lessening the PQ problems and supporting the functioning of sensitive loads are possible
because of Power Electronics and Advanced Control technologies [3]. The quality and reliability of
electric power distribution systems are reported to be improved by Custom Power Devices (CPDs).
Three major CPDs are D-statcom, DVR and UPQC [18]. One of the foremost custom power devices
that are competent of alleviating the consequence of power quality problems at the non linear load is
the UPQC [14] [15]. In addition to removal of harmonics, recompense for reactive power, load
current unbalance, source voltage sags, source voltage unbalance and power factor correction are
provided by UPQCs [16] [20]. In power distribution systems or industrial power systems, UPQC has
the outstanding potential to enhance the quality of voltage and current at the position of installation
[19].
        Generally, an UPQC is comprised of two voltage source inverters (VSIs) sharing with one
DC link capacitor. Here, the main problem is that the discharging time of DC link capacitor is very
high. To mitigate this problem, an enhanced ANFIS based UPQC is proposed and detailed in Section
4. Prior to that, a concise review about the recently available related works are given in Section 2.
Section 3 discusses problem formulation of the proposed technique. Section 5 discusses and analyzes
the results of the proposed controller and the Section 5 concludes the paper.

2. RELATED RECENT RESEARCHES: A BRIEF REVIEW

        Numerous research works already exist in the literature that compensate power quality
problem in power operating system. Some of them are reviewed here.
        A. Ananda Kumar and J. Srinivasa Rao [21] have proposed a novel method to improve the
power quality at the point of common coupling (PCC) for a 3-phase 4- wire DG system
using fuzzy logic control for grid interfacing inverter. The grid interfacing inverter is effectively
utilized for power conditioning. Their proposed method eliminates the additional power

                                                  75
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

conditioning equipment to improve power quality at PCC. They use fuzzy logic to handle rough
and unpredictable real world data. Their proposed method results show that, high accuracy of
tracking the DC-voltage reference, and strong robustness to load parameter variation.
         Miss Sobha rani Injeti and K. Ratna Raju [22] have presented a new compensation strategy
implemented using an UPQC type compensator. Their proposed compensation scheme enhances the
system power quality, exploiting fully DC–bus energy storage and active power sharing between
UPQC converters, features not present in DVR and D–Statcom compensators. The internal control
strategy is based on the management of active and reactive power in the series and shunt converters
of the UPQC, and the exchange of power between converters through UPQC DC–Link. They have
proved that their proposed algorithm was efficient and stability. Accurate, and robust in
comparing with the commonly used backward/forward sweep method for weakly meshed
networks.
         Lachman et al. [23] have proposed a Wavelet Transform technique for studying PQ
disturbances. They have discussed and deliberated Waveform distortion type of PQ disturbances.
Diverse kinds of signals have been created and diverse frequency bands have been acquired in the
calculation process. It has been observed that identification of PQ disturbances necessitates selection
of appropriate mother. It has been concluded that detection of PQ disturbances is made easier by the
use of approximate coefficients and scaling. Wavelet Transform has been proved to be an
appropriate tool for examining PQ disturbances, when time-frequency information is needed
simultaneously by means of computational results.
         Radhakrishnan et al. [24] have proposed a Fuzzy logic controller (FLC) for improving the
power quality indices of an AC–AC Conversion System as well as controlling its steady state and
transient state output voltage. The Fuzzy response verified by means of PI action has been proved to
be appropriate for utilization in Power Converters. The merits of the individual methods have been
illustrated by comparing the performance of three firing schemes. The applicability of Sinusoidal
Pulse Width Modulation (SPWM) over a wide range of applications has been demonstrated by
sufficiently emphasizing the fact that it inherits numerous exclusive merits. As Random PWM
(RPWM) considerably reduces the magnitude of harmonic components in addition to decreasing the
filtering necessities it adapts itself for high power applications by its ability of spread its harmonic
power. They have anticipated that such an analysis together with the advantages of the proposed FLC
will go a long distance in nourishing new applications for AC-AC conversion systems although the
option of a specific scheme could only be determined by specific needs.
         Darly Sukumar et al. [25] have discussed a fuzzy algorithm based new method for
attenuating the current harmonic contents in the output of an inverter. Ride-through capability amidst
voltage sags, decreased harmonics, better power factor and high reliability, reduced electromagnetic
interference noise, reduced common mode noise and increased output voltage range have been
provided by the inverter system that uses fuzzy controllers. A model of three-phase impedance
source inverter designed and managed based on the proposed considerations has been constructed
and a practicable test has been implemented. The improved effectiveness and suitability of these new
methods to reduce harmonics and increase the power quality have been verified from the practical
point of view. A compromise between cost and performance has been frequently required in the
realization because of the complexity of the algorithm. Variable-frequency DC-AC inverters, UPSs,
and AC drives have applications for the proposed optimizing strategies.

3. PROBLEM FORMULATION

       From the review of the research work shows that, an intelligent technique which is adopted in
various electrical and electronics applications. In the previous paper this intelligent technique is
being used for improving the power quality problem compensating performance of UPQC. This

                                                  76
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

technique combines the properties of both the neural network and the fuzzy logic which makes it
more robust and efficient. A neural network suggests the possibility of solving the problem of tuning.
Even though a neural network is capable of learning from the given data, the trained neural network
is generally understood as a black box. Neither it is possible to extract structural information from
the trained neural network nor can we integrate special information into the neural network in order
to simplify the learning procedure.
        Alternatively, a fuzzy logic controller is designed to work with the structured knowledge in
the form of rules and nearly everything in the fuzzy system remains highly transparent and easily
interpretable. However, there exists no formal framework for the choice of various design parameters
and generally the optimization of these parameters is done by trial and error method. Hence, a
combination of neural networks and fuzzy logic presents the possibility of solving tuning problems
and design difficulties of fuzzy logic. The resulting network will be more flexible and can be easily
recognized in the form of fuzzy logic control rules or semantics. This new approach combines the
well established advantages of both the methods and avoids the drawbacks of both. In this paper,
adaptive neural-fuzzy controller architecture is proposed, which is an improvement over the existing
Neuro fuzzy controllers.

3.1. Need for Advancement of Proposed Adaptive Technique
        As of now, various researches have been performed by researchers in the power quality
maintenance and enhancement sector. Some devices such as dynamic voltage restorer (DVR),
uninterruptible power supplies (UPS) and many others are used for maintaining the quality power
supply. But, these devices are capable of maintaining only the symmetrical or unsymmetrical power
supplies, so the power quality is not maintained at all time. To avoid these problems, an enhanced
Neuro fuzzy controller has been proposed [1] which is capable of maintaining the quality power
supply in the distribution network. But the system designed by [1] has a main drawback i.e. it can
only be applied to small networks and cannot work effectively in case of complex and wider
operating conditions.
        Hence to overcome this issue in this paper, we have proposed an intelligent technique called
the Adaptive Neuro Fuzzy Inference System (ANFIS) which is nothing but an advanced version of
the Nero fuzzy controller (NFC) or is a kind of neural network that is based on Takagi–Sugeno fuzzy
inference system. Since, it integrates both neural networks and fuzzy logic principles, it has a
potential to capture the benefits of both in a single framework. The fuzzy part of the ANFIS device
generally works with five steps which will be illustrated later. ANFIS also permits the combination
of numerical and linguistic data. Its inference system corresponds to a set of fuzzy IF–THEN rules
that have learning capability to approximate nonlinear functions. The generated fuzzy rules are then
trained by using the neural network and we get a desired output.

4. PROPOSED ADAPTIVE NEURO FUZZY INFERENCE (ANFIS) BASED UPQC

        UPQC is power electronics based power conditioning device, designed to compensate both
source current and load voltage imperfections. This device combines a shunt active filter together
with a series active filter in a back to back configuration, to mitigate any type of voltage and current
fluctuations and power factor correction in a power distribution network. UPQC is able to
compensate current harmonic reactive power, voltage distortions and control load flow but cannot
compensate voltage interruption because of not having sources. Hence in this paper we have
proposed a hybrid technique called the adaptive Neuro fuzzy inference system (ANFIS). By adding
an ANFIS device to the UPQC system the discharging time of the DC link capacitor is maintained at
a lower level. Accordingly the system performance is enhanced dramatically.


                                                  77
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

         The general structure of proposed ANFIS based UPQC consists of two inverters and it is
connected with a bias voltage generator. The inputs of the bias voltage generator are reference
voltage Vref and the calculated voltage Vcal, which is calculated from ANFIS. The Inverter II is
connected in series with an inductance that is denoted by LSLC. The purpose of the synchronous link
inductor is to generate a voltage with respect to PQ disturbance. The inverter I is connected in series
with a low pass filter and the purpose of the filter is to pass the low frequency component, and to
reduce the high frequency component of the specific voltage signal. Then, the output of low pass
filter is applied to the voltage injection transformer. Hence, the obtained output from injection
transformer maintains PQ in the operating system. The injected line voltage, voltage source, current
source and load current are denoted as Vinj, Vk, Ik, Iload . The Vinj can be determined as follows,

  2
Vinj = Vk2 − Vk2
         1     2
                                                                                            (1)
Vinj =   Vk2 − Vk2
           1     2
                                                                                            (2)
         nVdc
Vinj =                                                                                      (3)
         2 2

Where, Vk is the pre voltage variation, Vk is the post voltage variation,
            1                             2

‘n’ is the modulation index, Vdc is the normal rated DC voltage.

4.1. ANFIS based bias voltage generator
       The bias voltage generator is used for eliminating the high discharging time of the D.C link
capacitor. The ANFC is a hybrid technique which combines Fuzzy Inference System (FIS) and NN.
The fuzzy logic is operated based on fuzzy rule and NN is operated based on training data set. The
neural network training datasets are generated from the fuzzy rules and the error and change in error
voltage of the device is determined which is shown below,




                     Figure 1. Schematic diagram of proposed UPQC controller


                                                   78
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

                                        E ( k ) = Vdc ( ref ) − Vdc
                                        ∆ E = E ( k ) − E ( k − 1)
        Where, E (k-1) is the previous state error. The error voltage and change of error voltage are
calculated by using the above formula and the value are applied to the input of ANFIS. From the
output of NFC, the Vout is determined. The function of ANFIS is explained in t he below section.

4.2. Working Procedure of ANFIS
Step-1: At first the initialization of the input variables called the parameters is done which is in a
binary form and the input variables are fuzzy field.
Step-2: After input fuzzification, output fuzzification is done by applying fuzzy operators like AND,
OR operators.
Step-3: Membership functions are defined and are computed to track the given input/output data.
Step-4: The parameters associated with the membership function changes through the learning
process.
Step-5: Fuzzy rules are created basing on the input output relationship of the system.
Step-6: After creating rules, aggregation of various outputs is done and then the resulted functions
are defuzzyfied to get an optimal output.
Step-7: The obtained output is then trained by applying it to the Neural network through the back
propagation method.
Step-8: The error is minimized by performing various iterations in the Neural network and we get an
optimized output.

4.3. ANFIS Architecture
        One of the most popular hybrid methods called the adaptive Neuro-fuzzy method is nothing
but an advanced version of the Neuro-Fuzzy method. The ANFIS combines both the neural network
and the fuzzy logic controller methods. The basic structure of the type of fuzzy inference system
could be seen as a model that maps input characteristics to input membership functions. Then it maps
input membership function to rules and rules to a set of output characteristics. Finally it maps output
characteristics to output membership functions, and the output membership function to a single
valued output or a decision associated with the output. It has been considered only fixed membership
functions that were chosen arbitrarily. Fuzzy inference is applied only to modeling systems whose
rule structure is essentially predetermined by the user's interpretation of the characteristics of the
variables in the model.
ANFIS uses a combination of least squares estimation and back propagation for membership
function parameter estimation. The suggested ANFIS has several properties:
1. The output is zeroth order Sugeno-type system.
2. It has a single output, obtained using weighted average defuzzification. All output membership
functions are constant.
3. It has no rule sharing. Different rules do not share the same output membership function, namely
the number.
4. It has unity weight for each rule.
To explain the ANFIS architecture, two fuzzy if-then rules based on a first order Sugeno model are
considered.
Rule 1: If ( E is C1) and ( ∆E is D1) then (f1= p1 E + q1 ∆E + r1)
Rule 2: If ( E is C2) and ( ∆E is D2) then (f2= p2 E + q2 ∆E + r2)
Where, E and ∆E are the error and the change in error signals respectively, which are applied as
inputs to the ANFIS network, Ci and Di are the fuzzy sets, fi is the corrected and rectified signal
which we get as an output of the fuzzy region specified by the fuzzy rule, pi , qi and ri are the

                                                    79
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

design parameters that are determined during the training process. The bell shaped curves which are
having a maximum value of 1 and minimum value of 0 is shown below.




        fig.3. Shows the generalized ANFIS architecture which is formed by the combination of
adaptive neural network and adaptive fuzzy controller. The ANFIS operation is carried out by the
five layers and the processing of the five layers is explained below,
The network diagram and functioning of the ANFIS is explained through the five layers of the above
shown structure,
Layer 1: - The outputs of layer 1 are the fuzzy membership grade of the inputs which are all adaptive
nodes and are given by:
                                        Oi1 = µ Ci ( E ) , for i = 1, 2
                                                       Or
                                        1
                                      Oi = µ Di − 2 ( ∆ E ) , for i = 3, 4

        Where, E is the error and ∆E is the change in error, given as input to node-i, and Ci is the
linguistic label associated with this node function. In other words, O1i is the membership function of
Ci and it specifies the degree to which the given E satisfies the quantifier Ci. Usually µCi(E) is
chosen to be bell-shaped with a maximum value of 1 and a minimum value as 0, such as the
generalized bell function:

                       1
   µC ( E ) =                   2 ni
                                                                                               (4)
                    E − Ci 
                1+         
                    mi 

                                                 80
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

Layer 2: - In this layer every node is a circle node labeled ‘M’ which multiplies the incoming signals
and sends the product out. For instance,

                                  Oi2 = wi = µ C i ( E ) E µ D ( ∆E ), for i= 1,2
Each node output represents the firing strength of a rule. (In fact, other T-norm operators that
performs generalized AND can be used as the node function in this layer.)

Layer 3: - In the 3rd layer every node is a circle node labeled ‘N’ which represents normalization.
The ith node calculates the ratio of the ith rule’s firing strength to the sum of all rules firing strengths:

                                                wi
                                     Oi3 = w =       ,      for i= 1,2
                                             w1 + w2
For convenience, the output of this layer will be called as normalized firing strengths.

Layer 4: - This layer consists of square nodes I which encloses certain transfer functions in it.

                                       Oi4 = wf i = wi ( pi E + qi E + ri E )
where, wi is the output of layer 3, and { pi , qi , ri } is the parameter set. Parameters in this layer will be
referred to as consequent parameters.

Layer 5: - In the final layer a single circle node is present which is labeled as ‘Σ’ that computes the
overall output as the summation of all incoming signals, i.e.,

                                                  Oi5 = ∑ wi fi

          After evaluating the five layers of the ANFIS we get an error free signal which is our desired
output.
        By using the output of ANFIS, the DC-link voltage of UPQC is varied according to the
variation in load voltage. The ANFIS controller based regulating system, uses neural network for
optimizing the membership function of fuzzy controllers. The bias voltage generator is used for
eliminating the high discharging time of the D.C link capacitor. The inputs of the bias voltage
generator are taken as reference voltage Vref and the calculated voltage Vcal , which is calculated
from ANFIS. Initially, the load end and source end voltage of the system is applied to the neural
network. Then, the interference system is generated from the optimized output of neural network.
Accordingly, the performance of the ANFIS based UPQC device is evaluated and the D.C link
regulated voltage is determined. Then, the determined D.C link voltage is applied to the voltage
regulation system of UPQC and the PQ problem is compensated effectively.

5. RESULTS AND DISCUSSION

       The proposed ANFIS based UPQC controller is simulated in MATLAB platform. Then, the
performance of the proposed controller is tested with PQ problem. The testing performance of the
proposed controller is analyzed. The proposed ANFIS simulated diagram is illustrated below. The
parameters chosen for implementation are tabulated in Table I.




                                                        81
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

                          Table I: Parameters Chosen for implementation

                                 Parameters                     Values


                               DC Voltage (Vdc)                 ± 230V


                           Reference Voltage (Vref)             ± 230V


                              Error Voltage (e(n))        -230V to +230V

                         Change of Error voltage ( e)     -230V to +230V

       From the ANFIS based UPQC simulation model, the following performances are obtained.
The reference Voltage and Line Voltage at different instants, are determined. PQ affected the line
voltage has been enhanced using the proposed controller ANFIS, N.N and Fuzzy. However the
performance of the technique is analyzed by the process of ANFIS, NN-based controller and
then with FL. The reference voltage, line voltage with PQ problem at the defined time instants (T=
0.03,0.06,0.13,0.15,0.17 sec) and line voltage with enhanced PQ are illustrated in Fig 4,6,7,8,9, 10
and 11. Here, The PQ problems are occurring in NN, FLC and NFC at 0.03 seconds then clearing at
0.061, 0.062 and 0.061 seconds respectively. In the proposed controller, the PQ problems are
occurring at 0.03 seconds and clearing at 0.06 seconds. So, the total PQ problem duration is
0.03seconds.

                                                                                                 .




                                     Fig 4: Reference Voltage




                        Fig 5: Line voltage with PQ problem at T=0.03 sec



                                                  82
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME




                   Fig 6(a): Line voltage with PQ problem at T=0.03 sec




                   Fig 6(b): Line voltage with PQ problem at T=0.03 sec




                   Fig 6(c): Line voltage with PQ problem at T=0.03 sec




                   Fig 6(d): Line voltage with PQ problem at T=0.03 sec




                    Fig 7: Line voltage with PQ problem at T=0.06 sec

                                           83
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME




                       Fig 7(a): Line voltage with PQ problem at T=0.06 sec




                       Fig 7(b): Line voltage with PQ problem at T=0.06 sec




                       Fig 7(c): Line voltage with PQ problem at T=0.06 sec




                       Fig 7(d): Line voltage with PQ problem at T=0.06 sec

        From the above illustrations, the PQ problem arises at Time instant T=0.06 sec defined by
using ANFIS, NN and FLC methods. The PQ problems are occurring in NN, FLC and NFC then
clearing at 0.13, 0.14 and 0.14 respectively. In the proposed controller, PQ problems are clearing at
0.08 seconds. So, the total PQ problem duration is 0.02seconds.



                                                 84
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME




                    Fig 8: Line voltage with PQ problem at T=0.13 sec




                   Fig 8(a): Line voltage with PQ problem at T=0.13 sec




                   Fig 8(b): Line voltage with PQ problem at T=0.13 sec




                   Fig 8(c): Line voltage with PQ problem at T=0.13 sec




                                           85
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME




                      Fig 8(d): Line voltage with PQ problem at T=0.13 sec

         Here, the PQ problem arises at Time instant T=0.13 sec, and that are occurring in NN, FLC
and NFC then clearing at 0.16, 0.162 and 0.161 respectively. In the proposed controller, the PQ
problems are clearing at 0.16 seconds. So, the total PQ problem duration is 0.03seconds. Similarly,
all the different time instants, the proposed controller performance are calculated.




                        Fig 9: Line voltage with PQ problem at T=0.15 sec




                      Fig 9(a): Line voltage with PQ problem at T=0.15 sec




                      Fig 9(b): Line voltage with PQ problem at T=0.15 sec




                                                86
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME




                   Fig 9(c): Line voltage with PQ problem at T=0.15 sec




                   Fig 9(d): Line voltage with PQ problem at T=0.15 sec




                    Fig 10: Line voltage with PQ problem at T=0.17 sec




                   Fig 10(a): Line voltage with PQ problem at T=0.17 sec




                                            87
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME




                     Fig 10(b): Line voltage with PQ problem at T=0.17 sec




                     Fig 10(c): Line voltage with PQ problem at T=0.17 sec




                     Fig 10(d): Line voltage with PQ problem at T=0.17 sec

       From the above illustrations, the performance of the proposed controller has been
observed. So, the time taken to solve the PQ problem by ANFIS based UPQC system is very low
when compared to the NFC, FLC and NN. Therefore, the proposed controller is improved for solving
the PQ problems. The Root Mean Square (RMS) voltage (Vrms) is calculated as follows.
                                                     Vp
                                           V rms =
                                                      2
where, Vp is the peak voltage value.
 Then, PQ affected the line voltage has been enhanced using the proposed controller. Hence,
achieved RMS voltages of ANFIS, NN-based controller and FLC for different time instants are
tabulated in TABLE II.

                                               88
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

           Table II: RMS Voltages From ANFIS, NFC, NN-Based Controller And FLC

         Time instant in sec      RMS Voltage           RMS Voltage ( in Volts) after PQ
          at which the PQ        (in Volts) when                enhancement
            error occurs         PQ issue occurs      ANFIS   NFC         NN          FLC
                 0.03                  95                 77        28             49   63
                 0.06                  91                 88        31             53   81
                 0.13                  91                 81        53             53   56
                 0.15                  70             162.63        74             75   76
                 0.17                  88                 81        31             40   77

       The proposed controller enhances the PQ and brings the line RMS voltage to the
reference voltage level. But when the PQ problem is occurring in the defined time instants, the RMS
voltage gets decreased.

5.1 Performance Deviation
        The performance deviation of the proposed controller is analyzed with NFC, FLC and NN-
based controller. The performance deviation between ANFIS and NFC based controller can be
calculated as follows
                                                        ANFIS       NFC
                                                    (V rms − V rms ) * 100
                                  Deviation (%) =                ANFIS
                                                              V rms

        Similarly, the performance deviation can also be calculated between ANFIS and FLC,
NN for different instances of occurrence of PQ issues and that are chosen for implementing the
proposed controller are tabulated in TABLE III. The performance deviation of ANFIS controller
represented in fig 11. It can be deviates positively at a rate of 18.18% rather than FLC, 63.63%
rather than NFC and 36.36% rather than NN-based controller at T=0.03sec. And the time instant
(T=0.06 sec), the proposed controller deviates at a rate of 7.95% rather than FLC, 39.77% rather than
NN and 64.77% rather than NFC. Similarly, the performance deviation of the proposed technique
achieves a positive rate in solving the defined instants. It can be shown that, the proposed controller
can achieve a better performance of PQ issues compared with the FLC, NFC and NN.

                Table III. Performance deviation of ANFIS with NFC, NN and FLC
                      Time (in        Performance Deviation of ANFIS
                         sec)
                                      NFC            NN          FLC
                         0.03        63.63%        36.36%       18.18%
                          0.06         64.77%             39.77%         7.95%
                          0.13         34.57%             34.57%         30.86%
                          0.15         55.72%              52%               53%
                          0.17         61.73%             50.62%         4.94%


                                                     89
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME




         Fig 11: Performance analysis of ANFIS with NFC, FLC and NN-based controller

6. CONCLUSION

        In this paper, an ANFIS based UPQC controller was proposed for compensating the PQ
problem. The proposed controller was implemented and the results were evaluated with NFC, NN
and FLC methods. Here, the voltage sag PQ problems were considered for analyzing the
performance of the ANFIS controller. The voltage sag issues were represented at different time
instants of the line voltage. The proposed controller has compensated the PQ problem that is
specified with the waveform. The PQ problem compensated by ANFIS with 0.03seconds compared
with NN, NFC and FLC methods. The performance deviations were evaluated in the proposed
controller with the NFC, NN and FLC techniques. The performance deviation of the proposed
technique achieves a positive rate in solving all the defined instants of occurrences of PQ issues.
The comparative results showed that the proposed controller can achieve a better performance of PQ
issues compared with the FLC, NFC and NN based controllers.

REFERENCES

 [1]   S.Deepa and S.Rajapandian, "Voltage Sag Mitigation Using Dynamic Voltage Restorer
       System", International Journal of Computer and Electrical Engineering, Vol.2, No.5, pp.821-
       825, October 2010
 [2]   Thomas E. Grebe, "Application of Distribution System Capacitor Banks and Their Impact on
       Power Quality", IEEE Transaction on Industry Applications, Vol.32, No.3, pp.714-719, 1996
 [3]   S.Sadaiappan, P.Renuga and D.Kavitha, "Modeling and Simulation of Series Compensator to
       Mitigate Power Quality Problems", Vol.2, No.12, pp.7385-7394, 2010
 [4]   M.A. Hannan, A. Mohamed, A. Hussain and Majid al Dabbay, "Development of the Unified
       Series-Shunt Compensator for Power Quality Mitigation", American Journal of Applied
       Sciences, Vol.6, No.5, pp.978-986, 2009
 [5]   R.Valarmathi and A.Chilambuchelvan, "Harmonic Analysis of Distributed PV System Under
       High and Low Irradiance", International Journal of Engineering and Technology, Vol.2,
       No.3, pp.190-193, 2010


                                                90
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

 [6]    M.A. Beg, M.K. Khedkar, S.R. Paraskar and G.M. Dhole, "A Novel Clarke Wavelet
        Transform Method To Classify Power System Disturbances", International Journal on
        Technical and Physical Problems of Engineering(IJTPE), Vol.1, No.4, pp.27-31, 2010
 [7]    R. D. Patidar and S. P. Singh, "Active And Reactive Power Control And Quality
        Management In Dg-Grid Interfaced Systems", ARPN Journal of Engineering and Applied
        Sciences, Vol.4, No.3, pp.81-90, May 2009
 [8]    Mohammad Javad Dehghani, "Comparison of S-transform and Wavelet Transform in Power
        Quality Analysis", World Academy of Science, Engineering and Technology, Vol.50,
        pp.395-398, 2009
 [9]    Bulent Vural, Ali Kizil and Mehmet Uzunoglu, "A power quality monitoring system based on
        MATLAB server pages", Turk J Elec Eng & Comp Sci, Vol.18, No.2, pp.313-325, 2010
 [10]   T.Devaraju, V.C.Veera Reddy and M. Vijaya Kumar, "Role of custom power devices in
        Power Quality Enhancement: A Review", International Journal of Engineering Science and
        Technology, Vol.2, No.8, pp.3628-3634, 2010
 [11]   Kuldeep Sahay and Bharti Dwivedi, "Supercapacitors Energy Storage System for Power
        Quality Improvement: An Overview", J. Electrical Systems, Vol.10, No.10, pp.1-8, 2009
 [12]   S.S.Mortazavi, M.Razaz and E.Khavari, "Power Quality Improvement Using a Fuzzy Logic
        Control of a Series Active Filter", Journal of Theoretical and Applied Information
        Technology, Vol.19, No.1, pp.55-63, 2010
 [13]   Omer Gul, "An Assessment of Power Quality and Electricity Consumer’s Rights in
        Restructured Electricity Market in Turkey", Electrical Power Quality and Utilisation, Journal,
        Vol.XIV, No.2, pp.29-34, 2008
 [14]   Woo-Cheol Lee, "A New Control Scheme for Unified Power Quality Compensator-Q with
        Minimum Power Injection", Journal of Power Electronics, Vol.7, No.1, pp.72-80, January
        2007
 [15]   Othmane Abdelkhalek, Chellali Benachaiba, Brahim Gasbaoui and Abdelfattah Nasri, "Using
        of ANFIS And FIS Methods To Improve The UPQC Performance", International Journal of
        Engineering Science and Technology, Vol.2, No.12, pp.6889-6901, 2010
 [16]   C. Benachaiba, B. Ferdi, S. Dib and M. Rahli, "Impacts of Short-Circuit Power on Hysteresis
        Control of UPQC", European Journal of Scientific Research, Vol.37, No.4, pp.525-534, 2009
 [17]   T.Devaraju, V.C.Veera Reddy and M. Vijay Kumar, "Modeling and Simulation of Custom
        Power Devices to Mitigate Power Quality Problems", International Journal of Engineering
        Science and Technology, Vol.2, No.6, pp.1880-1885, 2010
 [18]   Yash Pal, A.Swarup and Bhim Singh, "Star-Hexagon Transformer Supported UPQC",
        International Journal of Electrical and Electronics Engineering, Vol.5, No.1, pp.63-68, 2011
 [19]   M. Davari, S. M. Aleemran, H. Nafisi, I. Salabeigi and G. B. Gharehpetian, "Modeling the
        combination of UPQC and photovoltaic arrays with Multi-Input Single-Output DC-DC
        converter", IEEE International Conference on Industrial Technology, 2009
 [20]   M. Hosseinpour, A. Yazdian, M. Mohamadian and J. Kazempour, "Design and Simulation of
        UPQC to Improve Power Quality and Transfer Wind Energy to Grid", Journal of Applied
        Sciences, Vol.8, No.21, pp.3770-3782, 2008
 [21]   A.Ananda Kumar and J.Srinivasa Rao, “Power Quality Improvement of Grid Interconnected
        3-phase 4-wire Distribution System using Fuzzy logic control”, International Journal of
        Engineering Research & Technology (IJERT), Vol. 1, Issue 4, June 2012.
 [22]   Miss Sobha rani Injeti and K. Ratna Raju, “Power Quality Improvement by UPQC device in
        Doubly Fed Induction Generator Wind Farm to Weak-Grid Connection”, International
        Journal of Scientific & Engineering Research, Vol.2, Issue 12, December-2011.



                                                  91
International Journal of Electrical Engineering and Technology (IJEET), ISSN 0976 –
6545(Print), ISSN 0976 – 6553(Online) Volume 4, Issue 4, July-August (2013), © IAEME

 [23] T.Lachman, A.P.Memon, T.R.Mohamad and Z.A.Memon, "Detection of Power Quality
      Disturbances Using Wavelet Transform Technique", International Journal for the
      Advancement of Science & Arts, Vol.1, No.1, pp.1-13, 2010
 [24] N.Radhakrishnan and M.Ramaswamy, "A Fuzzy Based Power Quality Enhancement Strategy
      For AC-AC Converters", International Journal of Engineering Science and Technology,
      Vol.2, No.4, pp.553-558, 2010
 [25] Darly Sukumar, Vanaja Ranjan and Benjamin Justus Rabi, "FLC Based Adjustable Speed
      Drives for Power Quality Enhancement", Serbian Journal of Electrical Engineering, Vol.7,
      No.2, pp.217-229, November 2010
 [26] Yash Pal, A. Swarup and Bhim Singh, "A control strategy based on UTT and I CosΦ theory
      of three-phase, fourwire UPQC for power quality improvement", International Journal of
      Engineering, Science and Technology, Vol.3, No.1, pp.30-40, 2011
 [27] M. Siahi, M. Najafi, M. Hoseynpoor and R. Ebrahimi, "Design and Simulation of UPQC to
      Improve Power Quality and Transfer Power of photovoltaic array to grid", Australian Journal
      of Basic and Applied Sciences, Vol.5, No.3, pp.662-673, 2011
 [28] Devendra Mittal, Om Prakash Mahela and Rohit Jain, “Detection and Analysis of Power
      Quality Disturbances under Faulty Conditions in Electrical Power System”, International
      Journal of Electrical Engineering & Technology (IJEET), Volume 4, Issue 2, 2013,
      pp. 25 - 36, ISSN Print : 0976-6545, ISSN Online: 0976-6553.
 [29] V. Niranjan and Ch. Das prakash, “Implementation of Wavelets with Multilayer and Modular
      Neural Network for the Compensation of Power Quality Disturbances”, International Journal
      of Electrical Engineering & Technology (IJEET), Volume 3, Issue 1, 2012, pp. 79 - 87,
      ISSN Print : 0976-6545, ISSN Online: 0976-6553.
 [30] N.Radhakrishnan and M.Ramaswamy, “Fuzzy Based Power Quality Enhancement Strategy
      For Three Phase AC-DC Converter”, International Journal of Electrical Engineering &
      Technology (IJEET), Volume 3, Issue 1, 2012, pp. 188 - 196, ISSN Print : 0976-6545,
      ISSN Online: 0976-6553.




                                              92

								
To top