 Open Access
 Total Downloads : 666
 Authors : G. Gowtham, A. Lakshmi Devi
 Paper ID : IJERTV4IS020483
 Volume & Issue : Volume 04, Issue 02 (February 2015)
 Published (First Online): 23022015
 ISSN (Online) : 22780181
 Publisher Name : IJERT
 License: This work is licensed under a Creative Commons Attribution 4.0 International License
Power Loss Reduction and Voltage Profile Improvement by DSTATCOM using PSO
G. Gowtham Prof. A. Lakshmi Devi
Electrical & Electronics Engineering Electrical & Electronics Engineering SV University SV University
Tirupati – 517502, India Tirupati 517502, India
AbstractThe objective of this paper is to reduce the power loss and to improve the voltage profile in radial distribution system. Fuzzy and particle swarm optimization algorithm is a twostage methodology used for placement and sizing of DSTATCOM in this paper. To show the effectiveness of the same a complete result analysis is carried out on 33 and 69 bus systems. Power loss and voltages are calculated for the four optimal locations based on priority.125%,150%,175% overloading cases are also considered in this paper. The result analysis shows that the two stage methodology effectively improves the voltages and reduces the power loss of the system.
KeywordsDSTATCOM, Fuzzy Approach, Particle Swarm optimization.
I. INTRODUCTION
Distribution networks find more importance in these days as it plays major role in power system planning and quality .The use of advanced equipments in distribution networks for quality improvement became necessary by the introduction of the deregulation in power systems. In distribution networks complete utilization of lines capacity is not possible for several reasons which lead to power flow limit decrease slower response time and increasing of power loss[10,11] . Modern techniques such as flexible AC transmission system (FACTS) works better in these regards. FACTS are initially developed for transmission system now these have been applied for distribution networks too. FACTS are of three types (i) series (ii) shunt (iii) combination of series and shunt. Shunt device DSTATCOM as a shunt connected voltage source converter is frequently used to compensate power quality. Under over loading and the voltage sag the load voltage of a particular bus can be regulated by the injection of compensating current into the system with the help of DSTATCOM[3]. A prototype design of DSTATCOM for voltage sag mitigation is presented for an unbalanced system [6]. A cascade loop control strategy to balance and regulate the voltage at a distribution bus using a DSTATCOM is proposed by [5]. various works [2,9] have been done on optimal location of STATCOM using various techniques such as and genetic algorithm (GA). Optimal placement and sizing of DSTACOM using immune algorithm[14]. Fuzzy approach
gives best optimal locations depending on the considered objectives and PSO technique iteratively optimize the sizes of the devices for the particular location. A MATLAB code is developed for the proposed approach and applied to IEEE 33 and 69 bus system and the results are tabulated.
The over loading cases 125%, 150%, 175% are considered in this paper.

MODELLING OF DSTATCOM
The STATCOM as a member of the FACTS devices is a regulating power utility which is connected to the power system in shunt mode. Once the STATCOM is used in the voltage level of distribution system is called Distribution DSTATCOM.
A DSTATCOM can work as synchronous voltage source with a variable magnitude and phase angle. Hence it is capable of controlling its bus voltage and correcting the power factor.
Usually DSTATCOM has the ability of injecting active and reactive power. Active power injection depends on the capacity of energy source. In this paper only DSTATCOM application for reactive power injection is considered and injection of active power is neglected.
Fig.1. A Typical DSTATCOM Connected To Bus

LOAD FLOW ANALYSIS
Generally used load flow analysis like Gauss Seidel , fast decoupled and Newton Raphson methods cannot be used to find the load flow in radial distribution systems because of high R/X ratio. Many special load flow analysis have been proposed in the literature[1,7].load flow analysis like load flow using conic programming [8], backward forward sweep based power flow analysis are also used. In this paper a direct approach for distribution system load flow solution [4] has been used.
The proposed algorithm is a novel but classic technique. The input data used in this algorithm is the conventional busbranch data which is used most. The aim of this algorithm is to develop a formulation, which solve the distribution load flow directly by taking the advantages of topological characteristics of distribution systems. It senses
that the time consuming methods like LU decomposition and forward/backward substitution of the Jacobian matrix substitution or the Y admittance matrix used in the Newton Raphson and Gauss implicit Z matrix algorithms are not necessary in the method used.
A bus injection to buscurrent matrix and a buscurrent to busvoltage matrix are the two developed matrices used here and by a simple matrix multiplication load flow solutions are obtained. The used method is robust and very efficient compared to conventional methods.

OPTIMAL LOCATIONS USING FUZZY APPROACH For Optimal Location of DSTATCOM on load buses
fuzzy approach is used in this paper [12,13]. Fuzzy logic is
developed by considering the following two objectives (i) power loss reduction (ii) maintaining voltage profile within the acceptable limits (0.9p.u 1.1p.u). Power loss reduction (PLI) and per unit nodal voltages (p.u) are taken as inputs to write fuzzy rules to determine the DSTATCOM placement suitability of each node. DSTATCOM can be placed on the nodes with highest suitability index.
i
i
LRi = P 1 P 2 (1)
Fig.3. membership function plot for p.u nodal voltage
Fig.4. membership function plot for DSTATCOM suitability index
By using the fuzzy rules from [13] the optimal placement of the DSTATCOM have been determined here in this paper.
Where i= 1 to number of load buses. LR = loss reduction.
i
P 1 = real power for normal load flow.
P 2 = real power for load flow by total compensation of

PARTICLE SWARM OPTIMIZATION
Particle Swarm Optimization (PSO) is proposed by James
I
reactive load at ith
node.
Kennedy and Russel C. Eberhart in 1995 which was inspired by fish schooling and bird flocks. It is a computational
The LR input is normalized by the following equation , so
that the values will fall between 0 to 1. Where the largest value will assign as 1 and the smallest as 0.
method that optimizes a problem by trying iteratively to improve the particular solution. Population of birds or fish is called as swarm. The particles selected from a particular
PLI = LR(I ) LR(min)
LR(max) LR(min)
(2)
range will move around in the search space according to a few formulae[13,15].
Let X and V are the position and velocity of the particles
The Fuzzy rules used in this paper are taken from (8f). DSTATCOM suitability index can be get by the output of the fuzzy. Maximum suitability index values are the optimal locations for DSTATCOM placement.
respectively. In a swarm by updating the position and velocity by the following formulas we will get personal best position (i.e. pbest) and global best position (i.e. gbest) the aim of the particles is to reach the gbest particle by using the formulas (3) and (4) .
V k+1 = WV k +C rand (pbest X ) + C rand (gbset
i i 1 1 i i 2 2 i
Xi ) (3)
X k+1 = X k + V k+1 (4)
i i i
Where,
Fig.2. membership function plot for power loss index (PLI)
V k = velocity if the particl i at Kth iteraton. W = Inertia weight parameter.
i
C1 = cognitive parameter. C2 = social parameter.
i
X k = particle position at Kth iteration.
rand1, rand2 = random numbers between 0 and 1.
Inertia weight can calculated by using the following equations for the better exploration of the search space .
W = wmax ((wmax wmin)*t)/T (5) wmax,wmin are the inertia weight factor constraints. t = current iteration count.
T = maximum number of iterations . Considered constraints are as follows
V min V V max (6)
Error = (max.fitness avg.fitness) (9)
If the calculated error is less than the specified tolerance then go to step 10.
i i i
X min X X max (7)
i i i

IMPLEMENTATION OF PROPOSED WORK
The PSO approach for solving optimal sizing of DSTATCOM to minimize the power loss and to improve the voltage profile takes the following steps:
step1: Get the inputs which are the line impedance and the bus data.
step2: Initially [nop x n] number of particles are generated where nop is the number of population and n is the number of DSTATCOOM devices.
step3: Generate initial [nop x n] number of velocities randomly between the limits. Iteration count is 1.
L
step3: load flow analysis is performed by placing all the n DSTSATCOM devices at the particular candidate locations and power losses P DSTATCOM are calculated. Same procedure is repeated for nop number of particles to find the total real power loses.
step4: for maximum loss reduction fitness function can be calculated by the following formula:
L
Fitness FA= PL P DSTATCOM
Step9 : the current iteration count is incremented, if the iteration coun not reaches maximum then go to step5.
Step10 : gbest fitness and the gbest particle gives maximum loss reduction and optimal sizes of DSTATCOM respectively.
6.1 Data used for PSO
nop = 100; C1=0.9; C2=0.9; wmax=0.8; wmin=0.1 T=100.

RESULT ANALYSIS
The proposed approach is used for optimal placement and sizing of DSTATCOM at the node which is having maximum power loss reduction and poor voltage profile are discussed in the following result analysis. The results compared with earlier published works[14].

Results of 33 bus system
Total real power loss in kW before DSTATCOM
Placement
Various Optimal locations
Size of
DSTATCOM
(kVAr)
Total real power loss in kW after DSTATCOM
Placement
30
1253.2
143.6445
202.7661
32
31
1032.2
1079.4
152.0194
150.0601
12
860.55
171.4469
Table.1 Results of 33bus system for various optimal locations
Where,
PL is total real loss before placement.
L
P DSTATCOM is the total real loss after placement of DSTATCOM.
Fitness with negative value is replaced with minimum and the respective particle position also assign with minimum from equation (7). Initially all the fitness values are copied to pbest fitness, maximum pbest fitness gives the gbest fitness. Which is a measure of of maximum loss reduction and the respective particles represents gbest particles.
Voltage levels of the considered 33 bus test system by placing DSTATCOM in various optimal locations obtained by fuzzy method have been plotted in the following Fig.5,
Fig.6, Fig.7, Fig.8 respectively
voltages of the system before and after placemnt of DSTATCOM at 30th bus
step5: using equations (3) (4) new velocities for all the particles within the limits are calculated and particle positions are updated respectively.
1
0.99
0.98
voltages (pu)
0.97
0.96
0.95
0.94
before placement
after placement
step6: after the updating of particles, load flow analysis is done and new fitness value is calculated using equation (6).
0.93
0.92
0.91
5 10 15
20 25 30 35
bus numbers
If the new fitness is greater than the pbest fitness then the
Fig.5.voltages before and after placing of DSTATCOM at 30th bus
respective particle is moved to the pbest particle.
Step7 : maximum pbest fitness gives the gbest fitness and the respective particle will be stored as gbest particle.
1
0.99
0.98
voltages (pu)
0.97
0.96
0.95
0.94
voltages of the system before and after placement of DSTATCOM at 32 bus
before placement
after placement
Step8 : maximum fitness and average fitness are calculated by using pbest fitness. Error is calculated using equation (9)
0.93
0.92
0.91
5 10 15
20 25 30 35
bus numbers
. Fig.6.voltages before and after placing of DSTATCOM at 32 bus
1
0.99
0.98
0.97
voltages before and after placement of DSTATCOM at bus 31
before placement after placement
Table.3 Results of 69bus system for various optimal locations
voltages (pu)
Total real
power loss in kW before DSTATCOM
Placement
Various
Optimal locations
Size of
DSTATCOM
(kVAr)
Total real
power loss in kW after DSTATCOM
Placement
225.0044
61
64
59
65
1330
1148.3
1383.8
982.2113
152.0446
160.0514
160.8188
169.0628
0.96
0.95
0.94
0.93
0.92
0.91
5 10 15 20 25 30 35
bus numbers
Fig.7.voltages before and after placing of DSTATCOM at 31 bus
before and after placement of DSTATCOM at the bus 12
1
before placement
after placement
0.99
0.98
0.97
voltage (p.u)
0.96
0.95
0.94
0.93
0.92
0.91
5 10 15 20 25 30 35
bus numbers
Voltage levels of the considered 69 bus test system by placing DSTATCOM in various optimal locations obtained by fuzzy method have been plotted in the following Fig.10,
Fig.8.voltages before and after placing of DSTATCOM at 12th bus
Fig.11, Fig.12, Fig.13 respectively.
Power losses have been calculated for the following test system at over loading conditions. The results have been tabulated as follows.
Table.2 Results of 33bus system for various overloading locations
1
0.99
0.98
0.97
voltage (pu)
0.96
0.95
voltages before and after placement of DSTATCOM at bus 61
before placement after placement
0.94
Loading condition
Losses without device
Location of device
PSO
Size of device (kVAr)
Losses with device
Normal
202.7661
30
1253.2
143.6445
125%
329.9988
30
1576.8
231.0839
150%
496.5653
30
1905.5
343.1695
175%
709.0218
29
2354.2
488.6277
0.93
0.92
0.91
0.9
10 20 30 40 50 60 70
bus numbers
Fig.10.voltages before and after placing of DSTATCOM at 61 bus
In the above table device represents DSTATCOM. Graphical representation of the voltages for the four loading conditions is as follows
1
0.99
0.98
0.97
voltage (pu)
0.96
0.95
0.94
0.93
voltages before and after placement of DSTATcom at bus 64
before placement after placement
0.92
0.91
0.9
10 20 30 40 50 60 70
bus number
Fig.11.voltages before and after placing of DSTATCOM at 64 bus
1
0.99
voltages before and after placement of DSTATCOM at bus 59
before placement after placement
0.98
0.97
voltages (pu)
0.96
0.95
0.94
Fig.9.voltages of various loading conditions
0.93
0.92

Results Of 69 Bus System
0.91
0.9
10 20 30 40 50 60 70
bus numbers
Results for various optimal locations to 69 bus test system according to their priorities obtained by fuzzy has been tabulated as follows,
Fig.12.voltages before and after placing of DSTATCOM at 59 bus
1
0.99
0.98
0.97
voltage (p.u)
0.96
0.95
0.94
0.93
0.92
0.91
voltages before and after placement of DSTATCOM at bus 65
before placement
after placement


CONCLUSION
The combination of fuzzy and PSO used as a twostage methodology in this paper to reduce the power loss and to improve the voltage profile in radial distribution system. Various optimal locations are obtained by the DSTATCOM suitability index from fuzzy. Optimal sizes for the respective locations are obtained by using PSO.The result analysis also
0.9
10 20 30 40 50 60 70
bus numbers
shows that the power loss of 125%, 150%, and 175% of normal loading is reduced and the voltage profile is
Fig.13.voltages before and after placing of DSTATCOM at 65 bus
Power losses have been calculated for the following test system at over loading conditions. The results have been tabulated as follows.
Table.4 Results of 69bus system for various overloading locations
Loading condition 
Losses without device 
Location of device 
PSO 

Size of device (kVAr) 
Losses with device 

Normal 
225.0044 
61 
1330 
152.0446 
125% 
369.0664 
61 
1675.1 
246.0443 
150% 
560.5439 
61 
2026.7 
367.7904 
175% 
809.4927 
59 
2517.8 
555.7028 
In the above table device represents DSTATCOM. Graphical representation of the voltages for the four loading conditions is as follows,
Fig.14.Voltages Of Various Loading Conditions
maintained within the limits. The voltages of all the loading conditions are also compared in this paper.
10. REFERENCES

Kresting WH, Mendive DL. An application of ladder network theory to the solution of three phase radial load flow problem. In: IEEE PES winter meeting; 1976. p. 760448.

Gerbex S, Cherkaouti R, Germond A. Optimal location of multitype FACTS devices in a power system by means of genetic algorithms. IEEE Trans Power Syst 2001;16(3):53744.

Sensarma PS, Padiyar KR, Ramanarayanan V. Analysis and performance evaluation of a distribution STATCOM for compensating voltage fluctuations. IEEE Trans Power Del 2001;16(2):25964.

Teng JenHao. A direct approach for distribution system load flow solutions. IEEE Trans Evol Comput 2003;18:8827.

Twining E, Newman MJ, Loh PC, Holmes DG. Voltage compensation in weak distribution networks using a DSTATCOM. In: IEEE conf PEDS, vol. 1; 2003. p. 17883.

Masdi H, Mariun N, Mahmud S, Mohamed A, Yusuf S. Design of a prototype DSTATCOM for voltage sag mitigation. In: IEEE conf power and energy; 2004. p. 616.

Wang Zhuding, Chen Fen, Li Jingui. Implementing transformer nodal admittance matrices into backward forward sweepbased power flow analysis for unbalanced radial distribution systems. IEEE Trans Power Syst 2004; 19:18316.

Jabr Rabih A. Radial distribution load flow using conic programming. IEEE Trans Power Syst 2006;21:14589.

Del Valle Y, Hernandez JC, Venayagamoorthy GK, Harley RG. Multiple STATCOM allocation and sizing using particle swarm optimization. In: IEEE conf power systems and exposition; 2006. p. 188491.

Somsai K, Kulworawanichpong T. Modeling, simulation and control of DSTATCOM using ATP/EMTP. In: 13th IEEE conf harmonics and quality of power; 2008. p. 14.

Sumpavakup C,Kulworawanichpong T. Distribution voltage regulation under threephase fault by using DSTATCOM, vol. 30; 2008. p.8559 [ISNN 1307 6884].

M. Damodar reddy and N.V.Vijay kumar, Optimal capacitor placement for loss reduction in distribution system using fuzzy and harmony search algorithm, ARPN journal of Engineering and applied sciences, vol.7, No.1, January 2012, ISSN 18196608.

M. Damodar reddy and K.Dhanunjaya Babu optimal placement of SVC using fuzzy and PSO algorithm International journal of engineering research and applications, vol.3, Issue 1, January February 2013 ,ISN: 22489622.

Seyed Abbas Taher and Seyed Ahmadreza Afsari optimal location and sizing of DSTATCOM in distribution systems by immune algorithm electrical power and energy systems 60(2014) 3444.

S.Devi and M.Geethanjali optimal location and sizing of Distribution static synchronous series compensator using particle warm optimization electrical power and energy systems 62(2014) 646653.