ECONOMIC LOAD DISPATCH OPTIMIZATION USING BAT

ALGORITHM WITH VARIOUS PERCENTAGES OF LOADS AT

DIFFERENT TIME INTERVALS

𝐏𝐫𝐞𝐞𝐭𝐢𝟏 , 𝐃𝐫. 𝐑. 𝐏𝐫𝐚𝐤𝐚𝐬𝐡𝟐 , 𝐑. 𝐀𝐧𝐠𝐮𝐫𝐚𝐣𝐚𝟑

1PG student, EEE Dept., DBIT
2Professor & Principal, EEE Dept., DBIT
3Associate Professor & HOD, EEE Dept., DBIT

Abstract- Nowadays the economic load dispatch is the vital issue in the electrical system due to the
greater requirement of the consumer. This issue can be optimized using improved technologies like BAT
algorithm, Genetic algorithm, Cuckoo algorithm and Intelligence water drop methods helps to upgrade
the economic dispatch issues. BAT algorithm was developed in 2010 and it is a meta-heuristic and
Nature- inspired algorithm which depends on echolocation behavior of micro-bats in searching their
prey. The strength of BAT algorithm is examined on IEEE30 bus system at various loads with different
time interval. The results represent the greater convergence capability and effectiveness of the BAT
algorithm.

Keywords- Bat algorithm (BA), Meta-heuristic, Loudness and Pulse rate of emission
I. INTRODUCTION
Economic dispatch is the process of optimization of operating cost is important in operational planning
issues. Nowadays organizing and utilization of power system is a demanding task due to its complexity
and to satisfy the consumer requirements for electrical energy with continuity of service and reliability.
The conventional approaches for solving the load dispatch issues are by applying non-linear
programming technologies. This approach minimizes a convex objective function over a convex set thus
insuring a single minimum. Meta-heuristic optimization is another method of resolving optimization
issues. These algorithms are commonly based on process in physics and biology. The meta-heuristic
algorithm is perfect for non-convex load dispatch issues as they do not affected from continuity. Bat
algorithm is the population based algorithm. This algorithm emulates the echolocation capability of
micro-bats which uses for recognizing and hunting the prey. The position of the Bat gives the possible
result for the load dispatch issues. Aim of the solution is represented by best position of BAT to its food.
BA and its constraints have been used to solve the load dispatch issues. Some well-liked nature-inspired
algorithms like Ant colony optimization (ACO) based on hunting behavior, Artificial bee colony (ABC)
based on behavior of honey bee, Cuckoo search based on the brooding behavior of cuckoo spices and
many more methods are used to optimize the load dispatch issues among them BAT algorithm has less
convergence time to operate hence it is very reliable method for power dispatch issue.

The Bat algorithm also applied for the non-smooth convex and non convex problems which has
unbalanced constraints which affect the economic operation of the system.

DOI : 10.23883/IJRTER.2017.3364.SDXHL 322

 International Journal of Recent Trends in Engineering & Research (IJRTER)
Volume 03, Issue 07; July - 2017 [ISSN: 2455-1457]

II. METHODOLOGY
2(a): Echolocation behavior of bats:
Echolocation is the use of sound waves and echoes to recognize where objects are in surrounding. To
Echolocate, bats send out sound waves from their mouth or nose. When the sound waves hit an object
they produce echoes.
fmin = fmin +( fmax − fmin ) * β
vit = vit−1 +( xit – x∗ )* fi
xit = xit−1 + vit
2(b): Loudness and pulse emission:
Micro bats release about 10-20 sound rapture per second. When searching for food this can varies up to
200 pulses a second as they are nearer to their prey. The loudness 𝐴𝑖 and the pulse emission rate 𝑟𝑖 are
reformed appropriately as the iteration proceeds. Once a BAT is recognized its food loudness reduces
and the rate of pulse emission factor varies
At+1
i = ∝ Ati
rit+1 = ri0 [1− exp(−γt)], whereα & 𝛽are the constants.
At the point of echolocation the micro-bats discharges string of short and high frequency noise impacts
after showing interest for echoes that rebounds back from the nearer objects. According to this echo the
Bats determines the object dimension, shape, direction and movement. Also it recognizes the distance
for its food. Even as the Bats move nearer to its food, the rate of pulse emission will increase as high as
200 pulses per 2d.

Fig 1: BAT sonar

The use of sonar (Sound Navigation And Ranging) with exceptional morphological and physiological
modification allows bats to detect with sound. Most bats produce echolocation sounds by contracting
their larynx. Echolocation calls are usually ultrasonic ranging in frequency from 20 to 200 kHz, whereas
human hearing will be more than 20kHz. In terms of loudness, bats emit calls as low as 50dB as high as
120dB, which is louder than smoke detector 10cm from our ear. That sound damages to human ear. The
ears and brain cells in bats are mainly attuned to the frequencies of sound they discharge and the echoes
that result. The exterior formation of bat’s ears also plays a special role in observing echoes.
Echolocation is highly technical and interesting strategy.

@IJRTER-2017, All Rights Reserved 323

 International Journal of Recent Trends in Engineering & Research (IJRTER)
Volume 03, Issue 07; July - 2017 [ISSN: 2455-1457]

2(c): Flow chart of BAT algorithm:

Fig 2: Flowchart of BAT algorithm

Description:
 Initiate the Bat population
 Define the system frequency and velocity
 Initiate the rate of pulse emission and the loudness
 Define the number of iterations wants to generate
 Find the best fitness function and update it
 If the rand value is less than rate of pulse emission find new best solution and update the velocities
and the current best solution
 If the rand value is less than loudness and the new best fitness function is greater than previous
function, accept the new best solution and update it
 Upgrade the bat position and the current best solution

@IJRTER-2017, All Rights Reserved 324

 International Journal of Recent Trends in Engineering & Research (IJRTER)
Volume 03, Issue 07; July - 2017 [ISSN: 2455-1457]

III. RESULTS AND DISCUSSION

The standard IEEE30 bus system is considered for the study, the generation cost and power losses has
illustrated in the table 1 at different % of load with time intervals. The graphs are obtained using
MATLAB coding.

Time Load Generation Power

Delay in (%) cost in loss
in hour ($/hr)
1 20 19607 204.39
2 60 7111.6 80.50
3 80 4657.2 52.03
4 100 662 9.18
5 40 29177 697.7
6 20 17832 188.54

Table 1: IEEE30 bus system at various % of load with time interval

Generation cost of 3rd hour with 80% load

Generation cost of 1st hour with 20% load

Generation cost of 2nd hour with 60% load Generation cost of 4th hour with 100% load

@IJRTER-2017, All Rights Reserved 325

 International Journal of Recent Trends in Engineering & Research (IJRTER)
Volume 03, Issue 07; July - 2017 [ISSN: 2455-1457]

Generation cost of 5th hour with 40% load Generation cost of 6th hour with 20% load

The above figures represent the variation of generation at different percentage of load at different time
interval.

IV. CONCLUSION
The convergence time is less for the Bat algorithm as compared to other methods. As shown in the table
1 the generating cost and total loss of the system is shown and the graphs are shown for each hour with
various percentages of loads using MATLAB coding. The BAT algorithm has less convergence time for
computations with moderate fuel cost. The results represent efficient and reliable behavior for solving
economic dispatch issues. Bat algorithm need less memory as the computations are less. Bat algorithm
works for multi-model systems which are having complex constraints and unbalanced parameters which
affect the economy of power dispatch. The improved BAT algorithm works for the non-convex dynamic
problems which can reduce the economy dispatch problem and optimize the generating cost of the
system.

REFERENCES
1. S.Vijayaraj and Dr. R. Santhi, “Economic dispatch with prohibited zones using improved Bat algorithm”, Vol 5, Issue 11, November
2016.
2. P. Sudheer Kumar Reddy and P. Anil Kumar, “Application of Bat algorithm for optimal power dispatch”, Vol 2, February 2015.
3. Farhad Namdari & Reza sedaghati, “Non-smooth economic dispatch solution by using enhanced Bat-Inspired optimization
algorithm”, Vol8, No 4, 2014.
4. Yamina Ahlem GHERBI and Hamid BOUZEBOUDJA, “Economic dispatch problem using Bat algorithm”, Issue 24, June 2014.
5. Aadil Latif and Peter Palensky, “Economic dispatch using modified bat algorithm”, July 2014.
6. Yubao Liu, “Improved bat algorithm for reliability-redundancy allocation problems”, Vol 10, No. 2, 2016.
7. Hegazy Zaher and Mahmoud EI-Sherbieny, “International journal of advanced research in computer science and software
engineering”, Vol 7, Issue 2, February 2017.
8. Gajendra sahu & Kuldeep swarnkar, “Optimization of economic load dispatch problem using genetic algorithm”, Volume 3, Issue
10, October 2014.
9. Nagendra singh & Yogendra kumar, “Multi objective economic load dispatch problem solved by new PSO”, Volume 2015.
10. Aadil Latif & Peter Palensky, “Economic dispatch using modified bat algorithm”, July 2014.
11. Anand R & A Azeezur Rahman, “Solution of unit commitment problem using BAT algorithm”, Volume 1, Issue 2, May 2014.
12. S. Jaisiva & A. Bharaneetharan. “Active power loss minimization using differential evolutionary based bat algorithm”, Volume 3,
Issue 09, February 2017.
13. Fatma Sayed Moustafa & N. M. Badra, “Solution of economic load dispatch using recent swarm-based meta-heuristic algorithms: A
survey”, 2016.
14. B. Vijay Kumar & N. V. Srikanth, “Bat algorithm and firefly algorithm for improving dynamic stability of power systems using
UPFC”, Volume 8, March 2016.

@IJRTER-2017, All Rights Reserved 326