IEEE - 31661

SOLUTION OF FIREFLY ALGORITHM FOR

THE ECONOMIC THEMAL POWER
DISPATCH WITH EMISSION CONSTRAINT
IN VARIOUS GENERATION PLANTS
THENMALAR.K 1

Dr.A.ALLIRANI 2

Assistant Professor, Dept of EEE

Vivekanandha College of Engineering for Women,
Namakkal (Dist.), India
thenmalark@gmail.com

Principal,
S.R.S College of Engineering & Technology,
Salem (Dist.), India
allirani2004@gmail.com

Abstract Economic load dispatch (ELD) is an important

optimization task in power system. It is the process of
allocating generation among the committed units such that
the constraints imposed are satisfied and the energy
requirements are minimized. There are three criteria in
solving the economic load dispatch problem. They are
minimizing the total generator operating cost, total
emission cost and scheduling the generator units. In this
paper Firefly Algorithm(FA) solution to economic dispatch
problem is very useful when addressing heavily
constrained optimization problem in terms of solution
accuracy. Results obtained from this technique clearly
demonstrate that the algorithm is more efficient in terms
of number of evolution to reach the global optimum point.
The result also shows that the solution method is practical
and valid for real time applications In this paper the
Firefly Algorithm(FA) solves economic load dispatch
(ELD) power system problem of three generator system,
six generator system with emission constraints and twelve
generator system with
introduced population-based
technique is utilized to solve the DED problem. A general
formulation of this algorithm is presented together with an
analytical mathematical modeling to solve this problem by
a single equivalent objective function. The results are
compared with those obtained by alternative techniques
proposed by the literature in order to show that it is
capable of yielding good optimal solutions with proper
selection of control parameters. The validity and quality of
the solution obtained Firefly Algorithm(FA) based
economic load dispatch method are checked and compared
with Artificial colony algorithm(ABC),Particle Swarm
Optimization Algorithm (PSO),simulated Annealing
Algorithm(SA)
Keywords: ABC-Artificial Bee Colony Algorithm, DEDDynamic Economic Dispatch, FA-Firefly Algorithm, PSOParticle Swarm Optimization Algorithm

I.

INTRODUCTION

thermal dispatch of generators. All the methods as well

known for the economic dispatch problem as a convex
optimization problem and it assumes the whole of the
unit operating limit (Pmin) and the maximum generation
limit (Pmax) is available for operation. This paper work
deals with the economic scheduling problem namely the
economic thermal power dispatch. In seeking the
solution for the economic dispatch problem (EDP) the
main aim is to operate a power system in such a way to
supply all the loads at the minimum cost. The solution
technique which is applied to the economic load
dispatch is Firefly Algorithm.The ultimate goal of power
plants is to meet the required load demand with the
lowest operating costs possible while taking into
consideration practical equality and inequality
constraints algorithms. Optimal operation of electric
power system networks is a challenging real-world
engineering problem. Those linked optimization
problems are the unit commitment, optimal power flow,
and economic dispatch scheduling.
The recently introduced meta-heuristic methods are
the artificial bee colony (ABC) algorithm, adapted
simulated annealing algorithm (SA),[14] firefly
algorithm (FA). These are a population-based technique
and inspired by the intelligent foraging behavior of the
honeybee swarm, birds schooling, fish behavior. The
DED problem was one of the real-world optimization
problems that has benefited from the development of the
meta-heuristic algorithms. An evolutionary programming
(EP) technique [8] in and is adopted to solve the DED
problem. Simulating annealing method (SA), quantum
evolutionary algorithm (QEA), Particle Swarm
Optimization Algorithm (PSO)
and Tabu search
approach (TS) have been also designated to solve the
DED problem in [12],[13], respectively.

Economic dispatch is one of the most important

problems to be solved in the operation of power system.
The traditional method such as lambda iteration method,
the base point, the participation factors method and
gradient method are well known for the economic

4th ICCCNT 2013

July 4 - 6, 2013, Tiruchengode, India

IEEE - 31661

-------- (4)
II.

OBJECTIVE FUNCTION

The objective function of the DED problem is to

minimize the operating fuels costs of committed
generating units to meet the load demand, subject to
equality and inequality constraints over a predetermined
dispatch period, the results practical usefulness will be
degraded if the units valve-point effects are neglected.
Consequently, there are two models to represent the
units valve-point effects in the literature. The first
represents the units valve in terms of prohibited
operating zones which are included as inequality
constraints. The second form represents the units valvepoint effects as a rectified sinusoid term which is
superimposed on the approximate quadratic fuel cost
function. The general mathematical form of the DED
problem follows: [1]
A.

Minimization of Fuel Cost

The problem of an Economic Load Dispatch (ELD)

is to find the optimal combination of power generation,
which minimizes the total fuel cost, under some
constraints [14]. The ELD Problem can be,
mathematically,
formulated
as
the
following
optimization problem:

------ (1)
PGi : the power generated by generator i (MW), and
n : the number of generators
B.

Therefore, to incorporate the constraints of units ramprate limits in the real power output limit constraints. The
modified units real power outputs are evaluated as
follows:

-------- (5)
Where Fcost : the total fuel cost ($/hr) ai,bi,ci : the fuel
cost coefficients of generator i
D.

Problem Formulation

1) Economic Dispatch Problem

The economic dispatch problem is defined as to
minimize
Fi = (aiPi2 + biPi+ci). Rs / hr.
n
Ft = (aiPi2 + biPi+ci). Rs / hr.
i=1
--------- (6)
Subject to
1. Pi,min Pi Pi,max
n
2. Pi = PD + PL
--------- (7)
i = 1 (The system demand constraint)

Equality Constraint

Integration of a renewable source (modifies the

equality constraints function to be as follows

2) Minimum NOX Emission dispatch

Ei = di Pi2 + ei Pi + fi , kg / hr.

Where pDt and pLt are the load demand and

systems loss at a time t respectively. The multiplier RS
is set to a permissible amount of active power injected
by RS, PRS is the forecasted real power from RS

The minimum NOx emission dispatch problem is

defined as to minimize
n
Ei = (di Pi2 + ei Pi + fi), kg / hr.
i=1
--------- (8)
Subject to

C. Inequality Constraint

1. The operating constraints - Pi,min Pi Pi,max

The inequality constraints of the DED problem are

the units ramp-rate limits, i.e., upper rate (URi) and
down rate (DRi), are considered as follows:

n
2. System demand constraint - Pi = PD + PL
i=1

-------- (2)

Minimum NOx algorithm is similar to the minimum cost

algorithm.
-------- (3)
Additional inequality constraints are the minimum and
maximum power output of each unit:

4th ICCCNT 2013

July 4 - 6, 2013, Tiruchengode, India

3) Combined Economic and Emission Dispatch

IEEE - 31661

A. Flow Chart of FIREFLY ALGORITHM

The Combined economic and emission dispatch

problem is defined as to minimize.

Start

n
n
i = Fi + h Ei , Rs / hr. --------- (9)
i=1
i=1

Initialize location of fireflies

Insert variable x into load flow

n
i= (aiPi2 + biPi+ci)+h (diPi2 + eiPi + fi)
i=1
--------- (10)

Run load flow

Subject to
The operating constraints Pi,min Pi Pi,max
n
System demand constraint Pi = PD + PL
i=1
Once the value of price penalty factor is known
the above equation can be rewritten in terms of known
coefficients and the unknown outputs of the generators.

Objective function evaluation

Ranking fireflies by their light

n
i = [(ai + hdi) Pi2 + (bi + hei) Pi + (bi + hfi) Rs / hr.
i=1
--------- (11)

Find the current best solution

Move all fireflies to the better

locations

h - Price penalty factor

III.

FIREFLY ALGORITHM

The development of FA is based on flashing behavior

of fireflies. There are about two thousand firefly species
where the flashes often unique for a particular species.
The flashing light is produced by a process of
bioluminescence where the exact functions of such
signaling systems are still on debating. Nevertheless,
two fundamental functions of such flashes are to attract
mating partners (communication) and to attract
potential. [9]
For simplicity, the following three ideal rules are
introduced in FA development. [12]
1)

all fireflies are unisex so that one firefly will be

attracted to other fireflies regardless of their
sex.

2)

attractiveness is proportional to their

brightness, thus for any two flashing fireflies,
the less brighter one will move towards the
brighter one.

3)

the brightness of a firefly is affected by the

landscape of the objective function.

For maximization problem, the brightness can

simply be proportional to the value of the objective or
fitness function.

Iteration Maximum

No

Yes
Run load flow program for final result
of optimal generation
Print results of the optimal dispatch, total
system loss and total generation cost
End

IV.

PARTICLE SWARM OPTIMIZATION

PSO simulates the behaviours of bird flocking.

Suppose the following scenario: a group of birds are
randomly searching food in an area. There is only one
piece of food in the area being searched.[2] All the birds
do not know where the food is. But they know how far
the food is in each iteration. So what's the best strategy to
find the food? The effective one is to follow the bird,
which is nearest to the food. PSO learned from the
scenario and used it to solve the optimization problems.
In PSO, each single solution is a "bird" in the search
space. We call it "particle". All of particles have fitness
values, which are evaluated by the fitness function to be
optimized, and have velocities, which direct the flying of
the particles. The particles fly through the problem space
by following the current optimum particles.
PSO is initialized with a group of random
particles (solutions) and then searches for optima by

4th ICCCNT 2013

July 4 - 6, 2013, Tiruchengode, India

IEEE - 31661

updating generations. In every iteration, each particle is

updated by following two "best" values. The first one is
the best solution (fitness) it has achieved so far. (The
fitness value is also stored.) This value is called pbest.
Another "best" value that is tracked by the particle
swarm optimizer is the best value, obtained so far by any
particle in the population. This best value is a global best
and called g-best. When a particle takes part of the
population as its topological neighbours, the best value is
a local best and is called p-best. After finding the two
best values, the particle updates its velocity and positions
with following equations .

ITERmax - maximum number of iterations

ITER - current number of iteration
A.

Flow
Chart
of
OPTIMIZATION

SWARM

Start
Initialize particles with random
position and velocity

Vi(u+1) = w * Vi(u) + C1 * rand( )* (pbesti - Pi(u)) +

vectors

C2 * rand( ) * (gbesti - Pi(u)) -------(12)

Pi(u+1) = Pi(u) + Vi(u+1)

PARTICLE

For each particle position (p)

evaluate the fitness

------ (13)

In the above equation,

The term rand( )* (pbesti - Pi(u)) is called
particle memory influence.
The term rand( ) * (gbesti - Pi(u)) is called
swarm influence.
Vi(u) is the velocity of ith particle at iteration
u must lie in the range

If fitness (p) is better than fitness o

(pbest) then pbest=p

Set best of pbest as g best

Vmin Vi Vmax

The parameter Vmax determines the resolution,

or fitness, with which regions are to be searched
between the present position and the target
position.

Update particle velocity and

Position

If Vmax is too high, particles may fly past good

solutions. If Vmin is too small, particles may not
explore sufficiently beyond local solutions.

In many experiences with PSO, Vmax was often

set at 10-20% of the dynamic range on each
dimension.

The constants C1and C2 pull each particle

towards pbest and gbest positions.

Low values allow particles to roam far from the

target regions before being tugged back. On the
other hand, high values result in abrupt
movement towards, or past, target regions.

The acceleration constants C1 and C2 are often

set to be 2.0 according to past experiences.

Suitable selection of inertia weight provides

a balance between global and local explorations,
thus requiring less iteration on average to find a
sufficiently optimal solution.

In general, the inertia weight w is set according to the

following equation,

------ (14)
Where w -is the inertia weighting factor
Wmax - maximum value of weighting factor
Wmin - minimum value of weighting factor

If g best is the
optimal solution

No

Yes

End
V.

EXPIREMENTS AND RESULTS

The power system economic dispatch problem

based on the concept of Firefly algorithm method has
been tested on 3-generator system, 6-generator system
and 12-generator system.Multiple generator limits and
total operating cost of the system is simulated in order to
evaluate the correctness as well as accuracy of this
method
A.

Three Unit System

The three generating units considered are

having different characteristic. Their cost function
characteristics are given by following equations
F1=0.00533P12+11.669P1+213 Rs/Hr
F2=0.00889P22+10.333P2+ 200 Rs/Hr
F3=0.00741P32+10.833P3+240 Rs/Hr

4th ICCCNT 2013

July 4 - 6, 2013, Tiruchengode, India

IEEE - 31661

According to the constraints considered in this

work among inequality constraints only active power
constraints are constraints are considered. There
operating limit of maximum and minimum power are
also different. The unit operating ranges are:
50 MW P1 200 MW
7.5 MW P2 150 MW
45 MW P3 180 MW

B.

The six unit generating units considered are

having different characteristic. Their cost function
characteristics are given by following equations.
Table 2: Fuel Cost Coefficient For Six Unit
System (PD=700mw)

The transmission line losses can be calculated by

knowing the loss coefficient. The Bmn loss coefficient
matrix is given by
Bmn =

0.0218 0.0017 0.0028

0.0093 0.0228 0.0017
0.0028 0.0093 0.0179

This is the example we have taken for the

testing this novel coding scheme on the three unit
system.
Here C1 = 1.99 and C2 = 1.99 Here the
maximum value of w is chosen 0.9 and minimum value
is chosen 0.4.the velocity limits are selected as V max=
0.5*Pmax and the minimum velocity is selected as V min=
-0.5*Pmin. There are 10 no of particles selected in the
population. For different value of C1 and C2 the cost
curve converges in the different region. So, the best
value is taken for the minimum cost of the problem.

P=
F=

a(p2)

b(p)

c
(Rs/hr)

pmin
(MW)

pmax
(MW)

F1

0.1524

38.539

756

10

125

F2

0.1058

46.159

451

10

150

F3

0.0280

40.396

1049

35

225

F4

0.0354

38.305

1243.

35

210

F5

0.0211

36.327

1658

130

325

F6

0.0179

38.270

1356

125

315

OBJE
CTIV
E
Econo
mic
Dispa
tch
Emiss
ion
Dispa
tch
Comb
ined
Econo
mic

The solution for ED problem of the three unit

system considered here is given below. The total
demand taken here is about 200 MW&700MW
1.0e+003 *
0.050000005801763
0.079584818649999
0.074549690507549
2.977211011227124
50.000005801763166
79.584818649999349
74.549690507549073
2.977211011227124e+003

Fuel
Cost
function

Table 3: Total Generation using different Techniques

(PD= 700 MW)

1) Simulation Results - Solution of ED Problem

out =

Six Unit System

P1

P2

P3

P4

P5

P6

41.37
485

19.56
447

122.0
292

95.87
92

230.2
597

215.0
321

86.20
00

84.83
09

110.9
136

111.0
101

161.7
126

162.6
996

67.02
26

64.07
49

115.9
380

119.4
393

174.9
152

177.2
328

Table 4: Fuel cost for different Techniques (PD= 700 MW)

Objective

Economic
dispatch

Emission
dispatch

Combined economic
and emission dispatch

FA

39299.56

40135.56

6123.45

ABC

40103.23

41533.65

6476.53

PSO

40392.66

42998.72

66239.72

Table 1: Fuel cost for different Techniques (PD= 700 MW)

Objective

Economic
dispatch

Emission
dispatch

Combined
economic and
emission dispatch

FA

20123.78

22234.45

32521.09

ABC

22693.16

23764.62

33987.90

PSO

23456.32

24038.33

34760.11

C.

Ten Unit System

The ten unit generating units considered are

having different characteristic. Their cost function
characteristics are given by following equations.

4th ICCCNT 2013

July 4 - 6, 2013, Tiruchengode, India

IEEE - 31661

Table 5: Fuel Cost Coefficient for 12 Unit

System(PD=1200MW)
Fuel
cost

a(p2)

b(p)

C
(Rs/
hr)

pmin
(MW)

pmax
(MW)

F1

0.0051

2.2034

15

12

73

F2

0.0040

1.9104

25

26

93

F3

0.0039

1.8518

40

42

143

F4

0.0038

1.6966

32

18

700

F5

0.0021

1.8015

29

30

93

F6

0.0026

1.5354

72

100

350

F7

0.0029

1.2643

49

100

248

F8

0.0015

1.2130

82

40

190

F9

0.0013

1.1954

105

70

590

F10

0.0014

1.1285

100

40

113

F11

0.0013

1.1954

105

70

190

F12

0.0014

1.1285

100

40

113

Table 6: Total Generation using different Techniques

(PD= 1200 MW)

Algorithm
P1
P2
P3
P4
P5
P6
P7
P8
P9
P10
P11
P12
Total
power
Fuel cost

FA
12
26
42
42.34
51.64
100
130
190
190
113
190
113

ABC
15
16
53
45.64
70.12
120
110
175
192
103
175
125

1199.98

1199.76

2.61E+003

2.7537
E+003

PSO
10
19.30
20
50
53.6
70
95
95
222
175
200
190
1199.99

VI.

An accurate solution is available on comparison

with other technique results. The validity of the proposed
method is demonstrated with the help of three standard
test systems. The Emission Constrained Economic
Dispatch (ECED) provides lower cost in emission. This
property is revealed for various load conditions.
The computational results of table (IV) consist
of twelve generator test system. It shows that FA fuel
cost function is less than ABC and PSO. The claims of
some of the recent reports provide near optimal solution
for large computationally intensive problem. Many of the
hybrid algorithms only help to improve the solution
accuracy. But this algorithm is very well defined and the
solution accuracy is excellent and converges to near
global minimum with less search account. It is high
efficiency than other methods. Thus, it obtains the
solution with high accuracy. An Firefly algorithm has
been developed for the economic thermal power dispatch
problem of electric generating units. The algorithm was
implemented and tested on 3-generator test system, 6generator test system and 12- generator test system.
Firefly Algorithm is an efficient tool for the economic
scheduling for generating units with the given generator
constraints and other data. The quality of the solution
shows that an improved Firefly Algorithm Search offers
a promising viable approach for solves the economic
thermal power dispatch problem.

2.652E+003

Table 4: Fuel cost for different Techniques

(PD= 300 MW,700 MW and 1200 MW)
Uni
t

Power
Demand
(mw)

Algorithm

Fuel cost
(RS/HR)

300

FA
ABC
PSO

1.32 E+003
1.47 E+003
1.57 E+003

Total
Power
(MW)
329
327
325

FA

3.215E+003

725.8721

700

12

1200

ABC
PSO
FA
ABC
PSO

3.5623 E+003
3.691E+003
2.61E+003
2.7537 E+003
2.652E+003

723.33
719.4354
1199.98
1199.76
1199.98

SUMMARY AND CONCLUSION

4th ICCCNT 2013

July 4 - 6, 2013, Tiruchengode, India

REFERENCES

IEEE - 31661

[1] A. J. Wood and B. F. Wollenberg, Power Generation Operation

and Control. New York: Wiley, 1996, pp. 3745
.
[2] M. Shahidehpour, H. Yamin, and Z. Li, Market Operation in
Elec- tric Power Systems. Forecasting, Scheduling, and Risk
Management. New York: Wiley, 1996, pp. 115160.
[3] Z. Li and M. Shahidehpour, Generation scheduling with thermal
stress constraints, IEEE Trans. Power Syst., vol. 18, no. 4, pp.
14021409, Nov. 2003.

[18] P.P.J. van den Bosch, "Optimal Dynamic Dispatch Owing to

Spinning- Reserve and Power-Rate Limits," Power Apparatus
and Systems, IEEE Transactions on, vol. PAS-104, no. 12, pp.
3395-3401, 1985.
[19] Economic Load Dispatch Using Firefly Algorithm K.
Sudhakara Reddy1, Dr. M. Damodar Reddy2 1-(M.tech
Student, Department of EEE, SV University, Tirupati 2(Associate professor, Department ofEEE, SV University,
Tirupati) July-August 2012

[4] R. A. Jabr, Optimal power flow using an extended conic

quadratic formulation, IEEE Trans. Power Syst., vol. 23, no. 3,
pp. 10001008, Aug. 2008.
[5] D. W. Ross and S. Kim, Dynamic economic dispatch of
generation, IEEE Trans. Power App. Syst., vol. PAS-99, no. 6,
pp. 20602068, Nov./Dec. 1980.
[6] Y. Al-Kalaani, Power generation scheduling algorithm using
dynamic programming, Nonlin. Anal., Theory, Meth., Appl., vol.
71, no. 12, pp. e641e650, Dec. 2009.
[7] Z.-L. Gaing, Particle swarm
optimization to solving the
economic dis-patch considering the generator constraints, IEEE
Trans. Power Syst., vol. 18, no. 3, pp. 11871195, Aug. 2003.
[8] N. Sinha, R. Chakrabarti, and P. K. Chattopadhyay, Evolutionary
pro- gramming techniques for economic load dispatch, IEEE
Trans. Evol. Computat., vol. 7, no. 1, pp. 8394, Feb. 2003.
[9] I. A. Farhat and M. E. El-Hawary,Optimization methods
applied for solving the short-term hydrothermal coordination
problem, Elect. Power Syst. Res., vol. 79, no. 9, pp. 13081320,
2009.
[10] I. G. Damousis, A. G. Bakirtzis, and P.S. Dokopoulos, Networkcon- strained economic dispatch using real-coded genetic
algorithm, IEEE Trans. Power Syst., vol. 18, no. 1, pp. 198205,
Feb. 2003.
[11] D. Karaboga and B. Basturk, "Artificial Bee Colony
(ABC) Optimization Algorithm for Solving Constrained
Optimization Problems," Foundations of Fuzzy Logic and Soft
Computing, pp. 789,2007
[12]Firefly Algorithm Technique for Solving Economic Dispatch
Problem M. H.
Sulaiman, M. W. Mustafa, Z. N. Zakaria, O.
Aliman, S. R. Abdul Rahim June2012

[13] M. Basu, Particle Swarm Optimization based goalattainment

method for dynamic economic emission dispatch, Electric
Power Components and Systems, vol.34, pp.1015-1025, 2006.
[14] S.Muralidharan, K.Srikrishna and S. Subramanian, Emission
constrained economic dispatch- A new recursive approach,
Electric Power Components and Systems, vol.34, no.3, pp.343353, 2006.
[15] B. K. Panigrahi, V. Ravikumar Pandi, and Sanjoy Das,
"Adaptive particle swarm optimization approach for static and
dynamic economic load dispatch," Energy Conversion and
Management, vol. 49, no. 6, pp. 1407-1415, 2008.
[16] T.E. Bechert and H.G. Kwatny, "On the
Optimal Dynamic
Dispatch of Real Power," Power Apparatus and Systems, IEEE
Transactions on, vol. PAS-91, no. 3, pp. 889-898, 1972.
[17] T.E. Bechert and Nanming Chen, "Area
Automatic
Generation Control by Multi-Pass Dynamic Programming,"
Power Apparatus and Systems, IEEE Transactions on, vol. 96,
no. 5, pp. 1460-1469, 1977.

4th ICCCNT 2013

July 4 - 6, 2013, Tiruchengode, India