International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 1

ISSN 2229-5518

Optimal Power Dispatch Incorporating UPFC Devices Using Foraging Algorithms Voltage Stability and Reliability Analysis

S.Jaganathan, S.Palaniswami

AbstractThe Refined Bacterial Foraging Optimization (RBFA) algorithm is biologically inspired computation technique which is inspired on foraging behavior of E-coli bacteria and its improved version of basic Bacterial Foraging Algorithm (BFA). This paper illustrated how a proposed algorithm different from other social behavior algorithm for solving optimization problems especially in multi-objective case and this proposed algorithm is applied to solve the Optimal Power Dispatch (OPD) problem with incorporating Unified Power Flow Controller (UPFC) and optimal location of UPFC based on stability and reliability analysis. This RBFA algorithm provides an enhanced optimal solution of all control variables and results are shown to be robust and improvement in convergence speed, simulations are performed with help of standard IEEE systems.

Index TermsAnt Colony Algorithm, Bacterial Foraging Optimization, FACTS Devices, Multi-Objective Optimization, Optimal Power

Dispatch, Refined Bacterial Foraging Optimization,, Reliability Analysis, Stability Analysis.

1 INTRODUCTION

—————————— ——————————
HIS paper illustrates and address the all power system problems with application Foraging Algorithm; the Bacterial Foraging Algorithm (BFA) introduced by Passino in 2002, and BFA is
based on behavior of E-coli bacteria.
The Bacterial Foraging Algorithms has been developed and ap- plied to solve various kinds of real valued problems, such has opti- mal non-linear dynamic system [23], PID controller design [20], optimal power flow [16], transmission network expansion planning [15], constrained optimization [13], Power System Stabilizer [22] and RFID network planning [18]. In this work, the improved version of bacterial foraging algorithm is presented and the proposed algo- rithm is called as Refined Bacterial Foraging Algorithm (RBFA).
The validity analysis of RBFA was carried out with various benchmark multimodal mathematical functions and results are com- pared with various social algorithms [13].The proposed RBFA is gives very much focus on towards the solution of chemotaxis behav- ior via search direction and search dimension. Generally the all op- timization algorithms are starts with random point only but end result shows time consuming process. Due to this reason and solution point of view, the BFA algorithm modified in to several stages and im- proved version of BFA proposed here.The proposed RBFA gives high potential to step length calculation and search dimension via
position updating process during chemotaxis behavior. The introduc-
repelling and attraction factor effects.Finally, in this paper the im- proved version of BFA algorithm is tested with multiple objectives power system problems or Multi Objective Optimization (MOO) problems.
This article presents the improved version of BFA (RBFA) algo-
rithm is applied to optimal power dispatch problems. The reliability and voltage profile analyzed with optimal placement of Unified Power Flow Control (UPFC) devices [2]-[9]. This particular OPD problem treated as Multi Objective Problem and this paper organized following manner, the following this section ‘2’ deals MOOPD prob- lem formulation, section ‘3’ illustrates review of Bacterial foraging Algorithms, section ‘4’ gives information about Refined BFA algo- rithms and algorithms for solving OPD problem with considering FACTS (UPFC) devices, section ‘5’ deliberate the results and discus- sions followed by comparison results and finally concluding of pro- posed work listed in section ‘6’.

2 PROBLEM FORMULATION

2.1 Objective Function

The multi-objective functions of OPD problem are mathemati- cally stated as follows
Minimize
tion of the attraction factor and velocity vector for improves the

F  

f f f f

(1)

1 2 3 4

————————————————
S.Jaganathan is currently pursuing PhD Degree in the Department of Electrical Engineering in Anna Universi- ty-Chennai. Email: jagananathangct@gmail.com
 S.Planiswami, Principal, Government College of Engi- neering, Bargur. His research interests are in power sys- tems and control systems. Email: joegct@gmail.com
speed of convergence and gives direction to global best according to the environments. The proposed algorithm also focused to process of
Subject equality and in-equality constraints are given below sec- tions.
The objective of OPD is to analysis reliability of power system under various operating conditions with results to identify the control variables for optimal operation of power system. The estimation of real power loss is to minimize the power loss via real power dispatch (fuel cost optimization) and reactive power planning with optimal placement of FACTS Devices (UPFC). The introduction of UPFC devices is to improve the stability, optimizing the real and reactive power dispatch and also improves the operating level of power sys-

IJSER © 2012

http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 2

ISSN 2229-5518

tem under critical operating condition.Finally, optimal placement and planning of reactive power requirement and identification critical lines and buses are connected with reliability analysis.
The objective functions are analysed and are given below:

Real Power Loss: This objective is to minimize the real power loss in transmission lines that can be expressed as:

-0.5XL<XTCSC<0.5XL
-200MVAR<QSVC<200MVAR (8)
Where
XL = original line reactance in p.u
XTCSC =reactance added to line
QSVC = reactive power injected at SVC placed bus in MVAR

f1   Nl G

 Vi2  V 2 – 2ViVjcos i – j

k 1

k  j

 (2)

Power Balance Constraints

Where Nl is the number of transmission lines; Gk is the conductance of the kth line; V < δ and V < δ are the voltages at the end buses i and
While solving the optimization problem, power balance equations are taken as equality constraints. The

i i j j

j of the kth line respectively.

Active Power Losses: P Pgi - Pl oad  0

power balance equations are given by,

+PL (9)

f2

k N gi

Ploss

K nE

g v2 v2

2vi v j cosij

(3)

Where
=Total power generation
=Total power demand
Where P is the total transmission active losses of the power system in MW; Pgi is the total active power generated in MW, Ploss the total transmission losses in the network to be calculated in MW and Pload is the total load of the system in MW.

Voltage profile optimization: Bus voltage is one of the most im- portant security and service quality indices. Considering only cost- based objectives in OPF problem may result in a feasible solution that has unattractive voltage profile. In this case, a two fold objective function is considered in order to minimize the fuel cost and improve voltage profile


PL =Losses in the transmission network

Pi= [Gikcos (-
(10)

Qi= [Giksin (- Where
Pi = Real power injected at bus i.
Qi = Reactive power injected at bus i.

= The phase angles at buses i and k

Ei, Ek = Voltage magnitudes at bus i and k

NG

f f w v 1.0P

Plim 2  NL V V lim 2 

NL Q

Qg lim 2 (4)

Real and Reactive Power Constraints

3  i

i P g1 g1

v i i Q

gi i

P min ≤ P
≤ P max , i=1,…., NG (11)

i  1

i NL i  1

i  1

Gi

min

Gi Gi

max

by minimizing the load bus voltage deviations from 1.0 per unit.

Average Voltage Deviation:

QGi
≤ QGi ≤ QGi
, i=1,…, NG (12)
Where

f  V

N V

 V N

(5)
P = Real power injected at bus i.

4 d i1

act des

gi

Qgi = Reactive power injected at bus i
Where Vd is the per unit (pu) average voltage difference; Vact is the actual voltage at bus bar i (p.u) and Vdes is the desired voltage at bus bar i (p.u).
Maximum Voltage Deviation:

Voltage constraints

Voltages VG are restricted by their lower and upper limits as fol- lows:
Vm =Max |Vact −Vdes|≥ 0 (6)

min

Gi

max

Gi Gi

, i=1,…., NG (13)

2.2 Constraints

Voltage Stability Constraints

Voltage Stability includes voltage stability constraints in the objective function and is given by [14]
Where NG is number of generators.

Transformer Constraints:

Transformer tap T settings are bounded as follows:

min i

≤ Ti ≤ Ti

max

, i=1… NT (14)

0 if 0.9<Vb<1.1
VS = 0.9-Vb if Vb>0.9
Where NT is number of Transformers.

Switchable VAR Sources Constraints:

Switchable VAR compensation QC is restricted by their limits as follows:
1.1-Vb if Vb>1.1 (7)

min

Ci

≤ Q max, i=1… NC (15)
Where
Vb= Voltage at bus b

UPFC Devices Constraints

Where NC is number of Switchable VAR sources.

Load Bus Voltages:

These include the constraints of voltages at load buses VL as fol- lows:
The UPFC device limit is given by,
VLi

Li Li

min ≤ V

≤ V max , i=1,…,NL (16)

IJSER © 2012

http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 3

ISSN 2229-5518

2.3 UPFC Devices

A Unified Power Flow Controller (UPFC) is a multifunctional
FACTS controller. The basic components of the UPFC are two
The Jacobian matrix J is:
Voltage Source Inverters (VSIs) sharing a common dc storage capacitor and connected to the power system through cou-
 P

J  

P

V    K L
(18)
pling transformer. The primary function of UPFC is power flow control and it’s because of series inverter. The series in- verter is controlled to inject the symmetrical three phase volt-
 Q


 
Q   

V
age system (Vse) in series with the line and there by enhance the power flow to desirable value, it is by controllable magni-
tude and phase angle series with the line to control active and reactive power flows in transmission line. The UPFC devices basically consist of two VSI connected back to back with inter- connecting DC storage capacitor. One VSI is connected to the system bus by a shunt transformer and other VSI is connected to the transmission line by a series transformer. The power balance between the series and shunt connected VSI is a pre- requisite to maintain a constant across the dc capacitor con- nected between the two VSIs [9]-[12].The Fig.1 clearly de- scribes controlling scheme in UPFC devices, how it’s connect- ed in critical lines, power injection to power.
Where K, L, M and N are the Jacobian sub-matrices, the ele-
ments of Jacobian matrix give the sensitivity analysis between power flow and bus voltage changes. The ∆P is incremental
change in bus real power, Q is incremental change in bus reac- tive power injection, ∆θ is incremental in bus voltage angle and ∆V is incremental change in bus voltage magnitude. In fact system voltage stability is affected by both P and Q. How- ever at each operating point we may keep P constant and evaluate voltage stability by considering the incremental rela- tionship between Q and V. Based on above considerations, equation (1), let ∆P=0. Then equation (1) become
 0 
 P

 
P

V   0 
(19)
dQ   Q
Q  dV

Vi i

q

Series transformer

V j qj

  
 
 
V

-1 ∆Q (20)

P, Q

Bus i Bus j

∆Q = JR ∆V or ∆V = JR
where

Shunt transformer

Converter

1

Converter

2

JR = [ JQV - JJPθ-1JPV] (21)
and J -1 is the reduced Jacobian matrix of the system for sensi-
Fig .1 Unified Power Flow Controller device.

2.4 Stability and Reliability Indices

The voltage stability and reliability problems can be effetively The voltage stability and reliability problems can be effetively analysed and solution is carried out with help of stability indi- ces, Voltage Stability Analysis, Fast line flow index, Voltage Stabilty Approch and Q-V Analysis. The detailed analysis of Voltage reliability analysis is listed in [5][7]. These methods examine the viability of the equilibrium point represented by a specified operating condition of the power system. The sensi- tivity analysis for voltage stability function indicates the eval- uation in a system condition and by solving the power flow equation of the network with specific load level and the load levels before the voltage instability.
This voltage indices calculation deals the relationship between voltage profile estimation and reactive power deviationchang- es at different buses using reduced Jacobian matrix.From the power flow equation
tivity. J -1 is also called sensitivity of Voltage stability analysis. A positive sensitivities represent stable operation; the smaller the sensitivity the more the stable the system. As stability de- crease, the magnitude of sensitivity becoming infinite at the maximum load ability limit. On the other hand, negative volt- age stability approach shows unstable operation. A small neg- ative sensitivity represents a very unstable operation. The magnitudes of sensitivities for different system conditions do not provide a direct measure of relative degree of stability. It is because of the non linear nature of the Q-V relationship and identifying the correct eigen values and eigen vectors for ana- lyzing the Voltage stability characteristics of the system in the reduced Jacobian matrix JR. and order of the reduction matrix is 1x1therefore the eigen values is equal to value of Jacobian reduction of the system[9]-[12].

3 BACTERIAL FORAGING ALGORITHMS

In 2000, Kevin.M.Passino [24] proposed the idea of Bacterial Forag- ing Methodology or Algorithm to solve the optimization problems. The base work of BFA was based on behavior of bacteria and very

 P

 dP   

P 

V   d

clearly states the foraging behavior of E-Coli Bacteria. In the BFA,
the optimization process consists of Chemotaxis, swarming (cell to cell signaling and foraging) reproduction and elimination- dispersal

     

(17)

dQ

Q

 

Q  dV

V 

activities. The optimization problem can be solved by in BFA, the

IJSER © 2012

http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 4

ISSN 2229-5518

way of search nutrients, evade noxious environments, the moving circumstance and the way of imitating the existence of the bacteria.
The solution set of BFA is based optimal sets of own. The posi- tion of each bacterium in the population is represented by [17][24].

P( j, k, l)  i j, k, li  1, 2,....S

Equations () describes the position of each bacteria, for ith is the
bacteria,jth chemotactic step, kth reproduction step and lth elimination-
dispersal event in θ co-ordinate axis in the solution surface.
The aim of BFA algorithm is to fit the best or minimum fitness function and the evaluating the fitness value of each bacterium is consist of following process cell-to-cell communication, its happens via attractant and repellent effect factor. This is given by the follow- ing equation
simulate the biological behavior of bacteria swarming. Reproduction comes from the concept of natural selection and the bacteria able to survive only if the bacteria adapted to their environment in best manner and transmit their genetic characters to succeeding genera- tions while those less adapted tend to be eliminated. Elimination – dispersal event selects parts of the bacteria to diminish and disperse into random position in the environment. The proposed method based on foraging behavior of bacteria and some modifications is made through from basic foraging algorithm. The modified algo- rithm named as Refined Bacterial Foraging Algorithm (RBFA) and RBFA is proposed in this paper. The performance of basic foraging algorithms is well suitable in single objective problems and static environments and step length of the basic BFA is a constant parame- ter which may guarantee good searching results for small optimiza-
tion problems. However it is applied to multiple objectives, dynamic

j i j, k, l ,j, k, l  i1 ji

,i j, k, l

environments and high dimensional problems and its gives poor per-

s  

s cc

p

i 2 

formance in convergence characteristics and also many times it trapped into local optima solutions. The process of search direction
and step length is important in multiple objectives and dynamic envi-
i  1dattract exp wattract m  1m m  

    

  

ronments. The proposed approach made some modification in step length; Chemotaxis behavior and adding velocity vector for improve

s  

p i

2 

speed of convergence and attaining global solution and also suitable

i  1hrepellant exp wrepellant m  1m m  

   

(22)

diversity for global search and its improves speed of convergence.

  

The finding of nutrient position and movement of each bacterium is
based on the tumble-run process, the every bacterium in the popula- tion will tumble once and the bacteria identify and have better fitness value after the tumble, the bacteria takes place run process with swim or swarm. The run length and step size important parameters in the BFA algorithm. The bacteria movement after the tumble-run pro- cess is given by in the equation

i ( j 1, k, l)  i j, k, l C i ‹ j

(23)
The evaluation of fitness and search process takes turns to execute until the life time of bacteria. After the cell to cell communication process, the reproduction schemes in follows and in this process the healthy bacteria identified. Finally, in this reproduction schemes the whole population will only keep half of the bacteria. Finally the elimination- dispersal process is takes place to eliminate the worst bacteria and by replacing the best bacteria [18][19][21].

4 REFINED BACTERIAL FORAGING ALGORITHM

The survival of species in any natural evolutionary process de-

Search direction and Chemotaxis process: The Chemotaxis behav-

ior is modeled by a tumble-run process that consists of a tumble step and several run steps. The tumble run process follows gradient searching principles, which means the bacteria’s position is updated in the run steps by the gradient information provided by the tumble step. In each bacteria step size, unit length, dimension and random direction coordination are much important, a search direction Wid (j) and a step length Lid (j) are calculated separately for each bacteria i, on each dimension d for each time step or tumble-run process or iteration j, Wid (j) =1 means the i-th bacteria goes towards the posi- tive direction of the coordinate axis on the dimension d, its indicates the follow tumble process and consecutive run process. Wid (j) = -1 means the bacteria goes towards the negative direction, its follow again tumble process and Wid (j) =0 means the bacteria stays at the current position; it means the bacteria reaches more nutrient position. When any one of bacteria finds best position, it should attract other bacteria so that they converge in that location. In this search direc- tion estimation leads to finding the distance of global bacteria from rest of others and the process convergence time minimized. While the distance known, it’s easy to attain the swarming process for new Chemotaxis stage. The current position ith the bacterium, in d dimen-
sional search space and jth tumble-run process is updated
pends upon their fitness criteria, which relies upon their food search-

Xid j 1 Xid j

 Lid j Wid j

(24)
ing and motile behavior. The law of evolution supports those species
who have better food searching ability and either eliminates or re- shapes those with poor search ability. The genes of those species who are stronger are propagated in the evolution chain since they posses ability to reproduce even better species in future generations.
So a clear understanding and modeling of foraging behavior in any
Once the direction and angle is decided by the tumble step, to reach the position it takes several steps (until the position of worst). The rotation angle Ф is related to the number of the dimensions, the number dimensions desired by objectives.
of the evolutionary species, leads to attempts better solution for any optimization. So, this is way here proposed the new algorithm.
The basic foraging algorithm consists of four steps are Chemo-

round d  1

(25)
taxis, swarming, reproduction, elimination and dispersal [1]. Chemo- taxis is the activity of bacteria gathering to nutrient-rich areas spon- taneously. A cell-to-cell communication mechanism is established to
Where, d- search dimension and π is the number of dimensions. Position updating process: In this process the step size calculation are initiated with help of position updating process. The basic meth-

IJSER © 2012

http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 5

ISSN 2229-5518

odology is adopted in BFA, In order to meet these criteria. The E-coli cells provides attraction signal to each other, so that they swarm or swim together and towards to the best location and swarming pattern based on cell-to-cell communication. But in this process shows slow converge only and in order to improve this process, the velocity fac- tor added with other bacteria so as to reach best bacteria position and this velocity vector added according to other bacteria position. The position updating process mainly focus on best bacteria and each dimension is given by

Where L i, j  Lid j Wid j

The J(i,j,k,l) is cost function of corresponding Xid (j,k,l).The fit-
ness value of the ith bacterium at Xid(j,k,l) is represented by jid (j,k,l).
In this paper the minimum fitness value jmin is defined as the global optimum.
In a run, consecutive steps of size L(i,j) is the same direction as the tumble is taken, in condition Xid (j+1,k,l) the cost function J (i,j+1, k,l) is better (lower) than J(i,j,k,l). In this condition, the fitness value of the ith bacterium in the rth step of the run Ns indicates the maxi- mum number of steps in a run and also r is smaller than Ns. This
Xid j Xid best  Xid
(26)
swimming operation is repeated as long as a lower cost is obtained until a maximum preset number of steps, Ns, is reached.
Where α is a factor to desire the strength of the bacterium’s at-
traction factor for improving the strength of the bacterium’s attrac- tion towards to the best position and for faster convergence., Xbest indicates the position of current best global solution updated after each function evaluation, and Xi is the position of the ith bacterium at the jth iteration after the tumble run process.
If a anyone bacteria reaches better position in order to direct the
In BFA the step length L is a constant and it ensures accuracy and
speed of the search. The size of the step length is dynamically ad- justed in the reproduction and elimination-dispersal process, which ensures the bacteria moving towards the global optimum quickly at the beginning, and converging to the global optimum accurately in the end. So the ith bacteria in dth dimension search space the step length controlling vector is given by
other bacteria to best position, the attraction signal is not enough to reach the best position. Therefore, in order to attract other bacteria
and also obtain the bacteria position by using equation
Where
L id  min L id µ, n
(29)
Xid j 1

Xid best  Xid  Vj

(27)
µ - step length vector, n = constant controlling the decreasing rate of
the step length. The value of ‘n’ is desired by according to position
So, in the process the velocity factor Vj is binding with position updating process and also is pilot to determine the best position ac- cording to the previous best positions. Update best bacteria position and add the velocity according to reach the best position, the velocity factor desired by depending upon according to position of best and other bacteria. The velocity factor Vj also gives direction to global best value or its used to identify global best value.

Step length: The basic BFA shows a step size is a constant parame- ter and it’s suitable to finding the solution in single and small optimi- zation problems. However, when applied to multiple objective prob- lems and dynamic environmental problems it shows poor perfor- mance. The basic BFA with fixed step size is failure in reaching the optimal point for the following reasons. The first case, if step size is

very high then the accuracy gets low although the bacterium reaches
updating process and Chemotaxis process.

5 ALGORITHM FOR REFINED BACTERIAL FORAGING

The multi-objective based OPD problem is can be solved by using
Refined BFA.
Read system data, here the standard IEEE30 bus system data [14] is considered. It contains load data, line data and generator data [14] with cost coefficients.
Randomly initialize the position of each bacterium in the domain, set the position and fitness value of the best bacterium. Initialize the parameters: S, p, Nc, NS, Nre, Nc, Ped, L(i), dattract, wattract, hrepellant and wrepellant
Randomly initialize the position of each bacterium in the domain, set
the position and fitness value of the best bacterium as Xid
the locality of optimum point quickly and it moves around the max-
ima for the remaining chemotactic steps. The second case, if the step size is very small, it takes many chemotactic steps to reach the opti- mum point and its lead to slow convergence and rate of convergence decreases. Meanwhile to reach the optimal point it takes convergence time and number of iterations increases. So, in order to reach the optimal point, speed of convergence and search ability the control- ling of step length is essential. In this paper the proposed RBFA fol-
and jmin (j, k, l) respectively.
FOR (each bacterium i=1: S)
FOR (chemotactic loop j=1: Nc)
FOR (dimensions=1: π)
FOR (reproduction loop k=1: Nre)
FOR (elimination-dispersal loop l=1: Ned)
Evaluate the cost function J(i,j,k,l)

b(j,k,l)

i

lowing consideration in determining step length process with help of
search dimensions, If deviation is very high then the step size must be increased and if the deviation is small in which case the bacterium is close to the optimum point the step size is to be reduced and also improve the performance the swim walk considered instead of the constant step.
For a each bacteria, the location of the ith bacterium at the jth chem-
Let Jlast = J(i,j,k,l) so that a lower cost could be found Calculate J (j,
k, l) and set it as Jlast; Tumble:
Search direction identified with help of identification process accord- ing to RBFA
For bacterium I, set Ji(j,r) as Jlast. A random vector (i), with each
element m(i), m=1,2….. p, a random number in the range [-1,1].
otactic step, rth reproduction step and l th elimination / dispersal event is represented by X (j,k,l) p. After a tumble, the location of the ith bacterium is represented by
Move:
Move to a random direction

'

 

by a unit walk, the

Xid j 1, r, l Xid j, r, l  L i, j j

(28)
new position is calculated by equation and start another chemotactic
step.

IJSER © 2012

http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 6

ISSN 2229-5518

Compute J(I,j+1,k,l) and use to compute Jcc (, P (j+1,k,l)) then use to find the new J(i,j+1,k,l).
Swim: let m = 0 (counter for swim length) While m< Ns (no climbing down too long) Let m=m+1

If J(I,j+1,k,l)<Jlast let Jlast = J(I,j+1,k,l) then take another step in the same direction and compute the new J(I,j+1,k,l).

Swim:
Update Xid(j+1, k, l). Recalculate Ji(j+1, k, l);
IF (Jcurrent< Jlast)
Step 4: After the load flow Solution, the multi-objective and mutli- disciplinary OPF are solved by using Refined BFA and all objectives are optimized by simultaneous optimization.

Step 5: Get the final optimal solution and quit the program, the above all procedure are clearly described in figure.2.

Start

Read System Data’s

WHILE (Ji
(j+1, k, l)<Ji (j+1, k, l)
and r<Ns)
Set Ji (j+1, k, l) as j

last;

Initialization

Run:
Update Xid (j+1, k, l)
Set new Ji
Swim:
(j+1, k, l) as J

current;

Load flow solution

Update Xid (j+1, k, l)
Recalculate Ji
END WHILE END IF
(j+1, k, l);

RBFA -OPD Optimal control settings

END FOR (bacterium)
Calculate Jmin (j+1, k, l) END FOR (chemotaxis) Sum:

Simultaneous optimiza- tion

Multi-objective function

health for the ith bacterium. Sort:

Evaluate the sum of the fitness value

Termination

Sort bacteria according to their health Ji
in ascending order.

Criteria

Split:
The bacteria with the highest Ji
values, computed by die while the
other Sr with the lowest values split and take the same location of their parents
Eliminate other then health bacteria are to be eliminated. Update:

Obtain the Global solution

Update the step length L (i,j). END FOR (reproduction) Disperse:
Disperse certain bacteria to random places in the optimiza- tion domain with probability Ped.
Update:

Secure voltage profile and all control variables

Update the step length L (id)
END FOR (elimination-dispersal) END
If it’s optimal solution is achieved means, the program will
stopped

5.1Computational Procedure:

Step 1: Read system data, here the standard IEEE30 bus system data [14] is considered. It contains load data, line data and gener- ator data [14] with cost coefficients. Input parameters of sys- tem, and specify the lower and upper boundaries of each var- iable.
Step 2: Initialize the number of bacteria and obtain the step length according foraging behavior.
Step 3: The load flow solution and control variables are initialized.
End

Fig.2: Flow chart for computational

Procedure of proposed algorithm

6 RESULTS AND DISCUSSION

The applicability and feasibility of the aforesaid tech- nique for practical applications have been tested with IEEE standard systems. The obtained results are compared with similar social behavior algorithms like ACO, Basic BFA and Modified BFA. The cases for taken for validity analysis and study comparison purpose the IEEE6 and IEEE30 bus taken.
The entire simulation is carried out with the help of

IJSER © 2012

http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 7

ISSN 2229-5518

MATLAB software and the challenging task of using the bio- inspired algorithms and tuning of FACTS devices (UPFC) are coded and all the tasks are obtained with MATLAB Packages. The fitness evaluation of the particles and the bacteria are car- ried out and evaluation of the proposed algorithm was tested with various multimodal function and analyzed in [].

6.1 IEEE 6 Bus System

In order demonstrate the effectiveness and robustness of the proposed technique the IEEE6 bus system is considered as first case and IEEE30 bus system is considered as second case. The solution of this problem with address of the minimization of real power loss, voltage deviation, voltage profile with sta- bility and reliability analysis and entire analysis carried out in two ways one is including UPFC devices and second way not considering the UPFC devices. The results of stability and reli- ability analysis and also optimized values are listed in Table.1. The results of proposed technique and comparison of results are tabulated in Table.2.

6.2 IEEE30 Bus System

The IEEE 30 bus system has 6 generator buses, 24 load buses and 41 transmission lines of which four branches are with tap settings transformers. The line data, bus data and other data are taken from []. The voltage magnitude limits at all buses are 0.95 p.u and the upper limits are 1.1 p.u. for all the PV buses and 1.05 p.u for all the PQ buses and the refer- ence bus. The lower and upper limits of the transformer tap- pings are 0.9 and 1.1 p.u respectively. The capacitor bank rat- ings are set as 0-15 MVAR.

TABLE 3

RESULTS IEEE 30 BUS SYSTEM

TABLE 1

RESULTS IEEE 6 BUS SYSTEM



The problem was handled as a multi-objective optimization problem where all the objectives are simultaneously optimized and problem analyzed with proposed technique and also simi- lar algorithms. The real power loss and voltage profile are considered as main function with proper identicfication of critical lines buses with help of stability indices [].

TABLE 2

COMPARSIONS OF RESULTS

TABLE 4

VOLTAGE PROFILE OF IEEE 30 BUS SYSTEM.



RBFA method -Voltage profile optimization
Voltage
Magnitude
1.0447
1.0391
1.0279
1.0253
1.0293
1.0321
1.0327
1.0272
1.0216
1.0189

1.0001

1.0257
1.0107
1.0059

0.9996

6.3 Comparison of Results

The validity of proposed algorithm and effectiveness of the solution are evaluated with help of the way of constraints handling, voltage stability and profile optimization and also Q-V analysis of the system [].

6.4 Handling of Constraints

The co-ordinate control of generator bus voltages, tap settings limits and reactive power limits are effectively optimized. The line

IJSER © 2012

http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 8

ISSN 2229-5518

losses reduced with optimal location of UPFC devices. The real power loss and reactive power loss greatly reduced with optimal location of UPFC devices with optimal control of control variables. The multi-objective OPD problem consists of equality and inequality constraints, the proposed approach is very efficient in handling of constraints and especially in in-equality constraints. The main focus of this problem is to obtain the optimal solution without violating any desired operating conditions.

When considering the bacteria search space and step length is very important, so as to achieve the good results the bacteria position is carried out in each step and velocity factor added to update the current position. The every possible solution the constraints should stay in the in their limits and here, it’s represented as power genera- tion boundary limits. Whenever the bacteria out of search space or bacteria swims out of the search space boundaries means the inequal- ity constraint is violated. In this proposed RBFA, the step length, search direction and position updated frequently, this will to find the better position of bacteria to achieve best solution and even if it moves out of the boundaries.

Fig.2. Voltage profile optimization of IEEE 30 bus system

6.5 Voltage Stability and Voltage Profile

The real and reactive power output in set by UPFC devices for during any contingency condition, therefore the optimal location of UPFC de- vices can maintain the voltage constant. The bus voltage deviation and average bus voltage deviation can be effectively minimized with help of optimal placement of FACTS devices. The voltage stability approach and stability indices are effectively identify the critical lines and critical buses with RBFA algorithm. So, the real power loss and voltage profile improvement are better. The results are tabulated in Table.1, Table.2, Table.3 and Table.4, the end results shows improvement in voltage sta- bility. The voltage profile optimization of IEEE30 bus system is clearly described in Fig.2.
The sensitivity analysis is better against critical conditions and load ability condition, So critical lines and buses are identified ac- cording to the series injection of UPFC devices and optimal location of UPFC devices with help of proposed algorithm. The impact of optimal location of UPFC is improvement in voltage stability and real power loss, finally the optimal dispatch is effectively analyzed
with addressing all the issues in the real time operation of power system.

7 CONCLUSION

This paper presents an improved version of bacterial foraging algorithm (RBFA) to solve multiobjective comprehensive model of optimal power dispatch problems with addressing the various objectives of optimal power dispatch problem, like real power loss, voltage stability, profile improvement and in addition to that optimize a various controls. The versatile op- timal power dispatch has been developed into modified envi- ronment in modern power systems with optimal operating and planning of resources. In feature include: optimal location of UPFC devices and voltage stability and reliability analysis for better voltage profile, reactive power planning, allocation of resources and reduction in line losses. The overall multi- objective solution effectively treated with proposed RBFA al- gorithm and this proposed method tested with various IEEE standard systems. The comparison of results using the pro- posed method is great improvements in speed, constraints handling, accuracy, optimal placement of UPFC devices and convergence in solving multi-objective and multi-constraint optimization problem.

ACKNOWLEDGMENT

The authors wish to thank Govt. Collge of Technology and RVS College of Engg and Tech for providing R&D lab and other facilities to carryout this work.

REFERENCES

[1] S. Jaganathan, Dr.S. Palaniswami, G. Maharaja vignesh, G., and R. Mithun- Raj, “Application of Multi-objective optimization in Reactive power plan- ning problem using a Ant colony Algorithm,” European Journal of Scien- tific Research, Issue 51, 2011, Vol 2,2011. ISSN 1450-216X Vol.51 No.2,

241-253.

[2] L. Gyugyi, “A unified power flow control concept for flexible AC transmis-

sion systems”, IEE Proc., Part-C, Vol.139, No.4, 1990, pp. 323-331.

[3] S. Tarakalyani, G. Tulsiram das, “Simulation of real and reactive power flow control with UPFC connected to a transmission line” Journal of Theo- retical and Applied Information Technology, 2008.

[4] Sheng-Huei Lee. Chia-Chi Chu, “Power Flow Models of Unified Power

Flow Controllers in Various Operating Modes" IEEE Trans. Power Elect,

2003, 0-7803-8110-6/03.

[5] J. Baskaran, V. Palanisamy, “Optimal location of FACTS device in a power system network considering power loss using genetic algorithm” EE-Pub on line journal, 2005.

[6] L.L Freris, A.M. Sasson, “Investigation of the Load Flow Problem”, Proc

IEE, Vol. 115, pp. 1459-1469

[7] A. Kazemi and B. Badrzadeh, “Modeling and Simulation of SVC and TCSC to Study Their Limits on Maximum Load ability Point”, 2004, vol. 26, no. 5, pp.381-388.

[8] A. Armbruster, M. Gosnell, B. McMillin, and M.L. Crow, “The maximum flow algorithm applied to the placement and distributed steady-state control of UPFCs,” in Proceedings of the 37th Annual North American Power Symposium, 2005, pp. 77–83.

IJSER © 2012

http://www.ijser.org

International Journal of Scientific & Engineering Research, Volume 3, Issue 11, November-2012 9

ISSN 2229-5518

[9] S. Singh, K. Verma, and H.Gupta “Optimal power flow control in open power market using unified power flow controller,” in Power Engineering Society Summer Meeting, 2001, vol. 3, pp. 1698–1703.

[10] N.G. Stagg, and H.A. El-Abiad, “Computer methods in power system analy-

sis’ (McGraw-Hill Inc., 1968)

[11] L. Abbate, M. Trovato, C. Becker, E. Handschin, “Advanced Steady-State

Models of UPFC for Power System Studies IEEE, 2002 pp .449-454.

[12] N.G. Stagg, H.A. El-Abiad, “Computer methods in power system analysis” (McGraw-Hill Inc., 1968).J. U. Duncombe, “Infrared navigation—Part I: An assessment of feasibility (Periodical style),” IEEE Trans. Electron Devices, vol. ED-11, pp. 34–39.

[13] S. Jaganathan, Dr.S. Palaniswami, G, “A New Refined Bacterial Foraging Algorithm for Multi- Disciplinary and Multi-Objective Problems,” Journal of Computational Intelligence in Bioinformatics, Vol.51 No.2, 113-131.

[14] S. Jaganathan, Dr.S. Palaniswami, K. Senthilkumaravel, and B. Rajesh, “Application of Multi- Objective Technique To Incorporate UPFC In Opti- mal Power Flow Using Modified Bacterial Foraging Technique”, Interna- tional Journal of Computer Applications (0975 – 8887), 2011, Vol.51 No.2,

241-253.)

[15] Jaganathan, Dr.S.Palaniswami, C.Sasikumar, R.Muthukumaran, 2010, “Multi Objective Optimization for Transmission Network Expansion Plan- ning using Modified Bacterial Foraging Technique,” International Journal of Computer Applications, vol 9 n.3, pp.28-34.

[16] S. Jaganathan, Dr.S. Palaniswami, G, “Solution of Multi – Objective Opti- mization Power System Problems Using Hybrid Algorithm,” Journal of

Asian Scientific Research, 2011, Vol.1 No.5, 265-270.

[17] Liu, Y., Passino, K.M., “Biomimicry of social foraging Bacteria for distrib- uted optimization: Models, Principles, and Emergent Behaviours,” J.Optimz Theory App., vol. 115, no. 3, pp. 603 – 628, 2002.

[18] Chen, H.N, Zhu, Y.L., Hu, K.Y, “Multi-colony Bacteria Foraging optimiza- tion with cell to cell communication for RFID network planning,” Applied soft computing, Vol. 10, no 2, pp. 539 – 547, 2010.

[19] Niu, B., Zhu, Y.L., He, X.X., Shen H., Wu, Q.H, “A life cycle model for simulating bacterial evolution,” Neurocomputing, Vol. 72, no. 1-3, pp. 142 –

148, 2008.

[20] Kim, D.H., Cho, J.H., “Adaptive tuning of PID controller for multivariable system using bacterial foraging based optimization,” In: Proc. 3rd Int, Atlan- tic web intelligence conf., Lodz, Poland, 2005, pp. 231 – 238.

[21] Dasgupta, S., Das, S., Abraham, A., and Biswas, A., “Adaptive computa- tional chemotaxis in bacterial foraging optimization: an analysis,” IEEE transactions on Evolutionary Computing, Vol. 13, pp. 919 – 941, 2009.

[22] Das, T.K., Venayagamoorthy G.K., Aliyu U.O., “Bio inspired Algorithms for the design of multiple optimal power flow system stabilizers: SPPSO and BFA,” IEEE transactions on Industry Applications, Vol. 44, no.5, pp.

1445 – 1457, 2008.

[23] Zhang, Y.D, Wu, L.N., “Efficient prediction of stock market indices using adaptive bacterial foraging optimization (ABFO) and BFO based tech- niques,” Expert systems with Applications, Vol 36, no.6, pp. 10097 –

10104, 2009.

[24] Kevin passion M, “Biomimicry of Bacterial Foraging for distribut-

ed optimization and control,” IEEE Control Systems Magazine,

2002, vol.22: pp.52-67.

IJSER © 2012

http://www.ijser.org