Optimization of Economic Dispatch Using Bacterial Foraging Optimization Algorithm

Melisa Astriyanti1, Siti Komsiyah2, Derwin Suhartono3

1Universitas Bina Nusantara, Jakarta Barat, Jakarta.

2Universitas Bina Nusantara, Jakarta Barat, Jakarta.

3Universitas Bina Nusantara, Jakarta Barat, Jakarta.

Email :

ABSTRACT

Economic Dispatch problem is one of electricity price problem. Economic Dispatch is one term to distribute electricity power generators in some active electricity system to minimize the cost of generators. The purpose of this study is to apply optimization method in Economic Dispatch problem to get optimal total price of thermal generators and then make web based application to simplify the calculation. Optimization method used is Bacterial Foraging Optimization Algorithm. Things that want to achieve is combination of electricity power generators and minimum total cost of generators and web based application to simplify the calculation. In the end, it can be concluded that this web based application is very useful in calculating the total cost of electricity power generators and the results is more optimal using Bacterial Foraging Optimization Algorithm method than the results of original data.

Keywords : Electricity Power, Economic Dispatch, Bacterial Foraging Optimization Algorithm.

INTRODUCTION

Electricity is one of individual needs in daily life. Electricity can’t be excluded in our life because almost there are no things that we can do without electricity. If there are no electricity in our life, human can’t do their works or fulfill their needs.

Technological and industrial developments and rapid population growth make demand for electricity continues to rise by years. Electricity could be produced from changes of natural energy or fuel. One of the largest manufacturers of electricity in Indonesia is PT PLN. PT PLN has subsidiary that is PT Indonesia Power. PT Indonesia Power has eight power generation units that will produce the electricity for consumer.

Each generation units has minimum and maximum power and has operating cost. Each generation units has unique charateristics and different energy. Thermal generators that using fuel like PLTD, has fuel cost that quite expensive when compared to operation cost of PLTA. Fuel cost is the largest cost that is almost 80% from total production cost.Because of that, combination in distribution of electricity power for each generators is a must. So minimum cost of generation units can be achieved.

Arrangement of combination in distribution power electricity can maximized work function of generation units if given an optimization methods. The amount of electricity power that has to be supplied from each generation units must be determined. So that total production cost can be minimized. Then, it can be said that distribution of electricity power is a cost function of generation units that also can be said as Economic Dispatch.

Economic Dispatch is one interesting problem and there are quite a lot of research about this problem using various kinds of optimization method. One of them is using Bacterial Foraging Optimization Algorithm method. Guozhong Wu researched about Economic Dispatch in hydro generation units using Bacterial Foraging Optimization Algorithm method. The conclusion is using Bacterial Foraging Optimization Algorithm method give an optimum results (Wu, 2012).

Other research using Bacterial Foraging Optimization Algorithm is done by R. Vijay. The result of this Bacterial Foraging Optimization Algorithm method is compared to other optimization method, that is Genetic Algorithm dan Particle Swarm Optimization. The conclusion is Bacterial Foraging Optimization Algorithm method give an better solution for Economic Dispatch problem than other methods (Vijay, 2012).

Bacterial Foraging Optimization Algorithm method also be done by Gautam Mahapatra dan Soumya Banerjee for solving simultaneous calculations problems. As for Genetic Algorithm, Bacterial Foraging Optimization Algorithm also used to resolve linear problems or non-linear problems and the results is better than conventional method (Mahapatra & Banerjee, 2013).

Other research is using hybrid Bacterial Foraging Algorithm-Particle Swarm Optimization with valve-point effects that result an optimal solution compared to Bacterial Foraging Algorithm method (Jayabarathi et al, 2012).

There is other research using Bacterial Foraging Algorithm method that has been upgraded that is Improved Bacterial Foraging Algorithm for solving economical operation of power system problems (EED) using wind power generation units. That EED be done with or without including wind power generation units to make a limitation and comparison in Bacterial Foraging Algorithm. The result is Improved Bacterial Foraging Algorithm method give a better result than usual Bacterial Foraging Algorithm (Farajianpour, 2012).

Other research about Economic Dispatch problem is using hybrid method of Bacterial Foraging Optimization Algorithm that is Fuzzy-Bacterial Foraging Optimization Algorithm. Unfortunately, there is no software that supports ease of calculation process (Fauzi, 2011).

Whereas for this research, Economic Dispatchproblem will be done using Bacterial Foraging Optimization Algorithm method and also a software that supports ease of calculation process will be made.

BacterialForaging Optimization Algorithm method has four foraging strategy that can work simultaneously with high sensitivity for optimization process. This method will be used for calculating the combination of electricity power of each generation units hoping that cost of generation units more optimal than the one that is not using this optimization method.

Implementation of this Bacterial Foraging Optimization Algorithm method for Economic Dispatch problem will be facilitated with a web based software that will helping complex calculation process. So the purpose of this research that is designing software for implementing Bacterial Foraging Optimization Algorithmmethod for Economic Dispatch problem to result a minimum cost of generation units will be achieved.

RESEARCH METHOD

In implementation, this study was made as follows :

1.Identification of the cost optimization problem, that is minimizing the cost to benefit as much as possible. Issues that will be examined is the problem concerning Economic Dispatch or scheduling electricity load problem. Literature study on electricity subject and Economic Dispatch is required to resolve these problems.

2.Selecting an optimization method which is the scope of Swarm Intelegencethat is Bacterial Foraging Optimization Algorithm method. Literature study on this method is very important so that later can be made an application based on this method to solve the Economic Dispatch problem.

3.Collecting the necessary data that will be processed to resolve these problems. Once the data obtained, the necessary simulation calculation can be done.

4.Coefficient of generation units cost, minimum and maximum generation units, and total load of geneation units are obtained from the company that is the object of research.

5.Calculation of electricity cost can be done with mathematical modeling of power cost function.

  1. Determine coefficient of heatrate (coefficient a,b,c) from generation units data with quadratic interpolation.
  2. Find the value of cost function coefficient (coefficient a,b,c) by multiplying coefficientof heatrate with energy prices (Rp/kcal).

Mathematical modeling of power cost function is as follows :

Where,

= total cost of generation units (Rp)

= cost function of generation units (Rp/hour)

= cost coefficient of generation units

= output of generation units (MW)

= total of generation units

= index ofdispatchable unit

Each generation units has a limitations that formulated as follows :

6.Next, process to Bacterial Foraging Optimization Algorithmmethod to find the value of each generation units.Here is the algorithm of Bacterial Foraging Optimization Algorithm method :

Picture 1 Flowchart ofBacterial Foraging Optimization Algorithm Method

7.Steps to be followed in this method is :

  1. Initialize the position of the bacteria in this case is random value of each power plant in accordance with the minimum to the maximum limit.

Picture 2 Initialization of Bacteria Position

Bacteria position define power quantity of each generation units that can be obtained from random numbers which is formulated as follows :

Where,

power quantity of generation units at bacteria

= minimum threshold of generation units

= maksimum threshold of generation units

= random number from 0 to 1

  1. Initialization the parameters ofBacterial Foraging Optimization Algorithm method, namely : Number of bacteria (), Number of generators (), Loop chemotaxis (), Loop swimming (), Loop reproduction (), Loop elimination and dispersal (), Probability of elimination and dispersal (), Bacteria step for swimming and tumbling (), dattract, wattract, hrepellant, wreppelant.
  2. Entering the course of this algorithm in four main steps Bacterial Foraging Optimization Algorithm method (Chemotaxis, Swarming, Reproduction, Elimination and Dispersal).
  3. Obtain the results that is combination of electricity load each generation units and minimum total cost.
  4. Check if the total load of each generation units meet the tolerance input desired.
  1. If the condition fulfilled, the results are displayed. Otherwise, it will continue to run up to meet the tolerance.

8.Designing a web-based application to calculate Economic Dispatch problem using Bacterial Foraging Optimization Algorithm method using HTML and PHP.

9.Comparing the calculation results from the application, which results is combination of electricity load of each generation units and minimum total cost, with original data that were obtained from company.

ANALYSIS AND RESULTS

Application Design

Picture 3 Use Case Diagram User

User can perform the functions contained in this web application, which are :

  1. Login which will restrict acess to perform calculating process in this web application so only authorized user from certain division in the company can use the calculating process.
  2. Use Process which is part of Process menu.After user logs in, user can use optimization computing application. User has to enter data input first.The input of data will be processed to result an ouput in the form of table and graphics.
  3. View Help serves as a guide that will guide the user in performing this calculation application.
  4. View About Method serves as information to the user regarding the optimization method used to process the data in this web application. Optimization method used is the method of Bacterial Foraging Optimization Algorithm.
  5. View Company Profile is an information about the company where case study was conducted.
  6. View About Creator is information about author of this application.
  7. Logout will logs out user from calculation process in this application. User will be directed to Home page.

Results of Data Process

Implementation of Bacterial Foraging Optimization Algorithm method for Economic Dispatch problem is found on a web-based software that has been made to simplify the calculation of the optimization method. Program tryout was conducted by doing five times experiments with same coefficient input value of equation cost (coefficient value of a, b, and c are same ), P Demand input value is 3321 MW, and Toleransi input value is 0,1.

  1. First Experiment

After the first trial run, obtained load of each generation units with a value of Error = 0,024165 dan Cost = Rp. 2.926,113128 per hour. Below is table of combination load generation units in first-mentioned expriments.

Table 1 Results of First Experiment

Generators / Power (MW)
PLTU A / 354,4491421
PLTU B / 359,2903176
PLTU D / 271,1874699
PLTU E / 492,0624132
PLTU F / 560,4367306
PLTU G / 457,3546876
PLTU H / 574,1432685
PLTU I / 252,0518051
Total / 3320,975835
  1. Second Experiment

After the second trial run, obtained load of each generation units with a value of Error = 0,01938 dan Cost = Rp. 2.933,88221 per hour. Below is table of combination load generation units in second-mentioned expriments.

Table 2Results of Second Experiment

Generators / Power (MW)
PLTU A / 367,36042
PLTU B / 243,09668
PLTU D / 359,58352
PLTU E / 410,56642
PLTU F / 567,75606
PLTU G / 563,17906
PLTU H / 603,95126
PLTU I / 205,487
Total / 3320,98062
  1. Third Expreriment

After the third trial run, obtained load of each generation units with a value of Error = 0,038687 dan Cost = Rp. 2.922,32356 per hour. Below is table of combination load generation units in third-mentioned expriments.

Table 3Results of Third Experiment

Generators / Power (MW)
PLTU A / 355,122748
PLTU B / 357,1216771
PLTU D / 356,8317946
PLTU E / 561,2485205
PLTU F / 492,2541262
PLTU G / 368,5028966
PLTU H / 574,2562657
PLTU I / 255,7007016
Total / 3321,038687
  1. Fourth Experiment

After the fourth trial run, obtained load of each generation units with a value of Error = 0,077857 dan Cost = Rp. 2.924,32447 per hour. Below is table of combination load generation units in fourth-mentioned expriments.

Table 4Results of Fourth Experiment

Generators / Power (MW)
PLTU A / 346,3869069
PLTU B / 287,5598931
PLTU D / 354,4203968
PLTU E / 514,9361055
PLTU F / 552,2352366
PLTU G / 515,3556477
PLTU H / 515,9109487
PLTU I / 234,27272223
Total / 3321,077858
  1. Fifth Expreiment

After the fifth trial run, obtained load of each generation units with a value of Error = 0,038687 dan Cost = Rp. 2.922,32356 per hour. Below is table of combination load generation units in fifth-mentioned expriments.

Table 5Results of Fifth Experiment

Generators / Power (MW)
PLTU A / 340,8370524
PLTU B / 341,0677507
PLTU D / 368,1866421
PLTU E / 431,9512944
PLTU F / 564,2833576
PLTU G / 456,7201545
PLTU H / 616,8870935
PLTU I / 201,1608698
Total / 3321,094215

Comparison Results

The table below represents the cost of original data to meet the load demand of 3321 MW.

Table 6 Generators Cost

Generators / Power (MW) / Cost (Rp/hour)
PLTU A / 372 / 351,0814846
PLTU B / 372 / 351,0814846
PLTU D / 372 / 351,0814846
PLTU E / 575 / 344,1076797
PLTU F / 575 / 344,1076797
PLTU G / 575 / 344,1076797
PLTU H / 200 / 401,262627
PLTU I / 280 / 465,263699
Total / 3321 / 2952,263699

While in five times experiments of program simulations, the results is on the table below :

Table 7 Average Results of Program Simulations

Average / Units
PLTU A / 352,8312453 / MW
PLTU B / 317,6272637 / MW
PLTU D / 342,0419647 / MW
PLTU E / 482,1529507 / MW
PLTU F / 547,3931022 / MW
PLTU G / 472,2224893 / MW
PLTU H / 577,0297673 / MW
PLTU I / 229,7346198 / MW
Total Power / 3321,033443 / MW
Total Cost / 2927,620782 / Rp/hour

Below is comparison table of original data with average of five times expriments that is generated by application using Bacterial Foraging Optimization Algortihm :

Table 8 Comparison Original Data and Average Results of Program Simulation

Original Data / BFOA Method
PLTU A / 372 / 352,8312453
PLTU B / 372 / 317,6272637
PLTU D / 372 / 342,0419647
PLTU E / 575 / 482,1529507
PLTU F / 575 / 547,3931022
PLTU G / 575 / 472,2224893
PLTU H / 200 / 577,0297673
PLTU I / 280 / 229,7346198
Total Power / 3321 / 3321,033443
Total Cost / 2952,263699 / 2927,620782

Based on comparison of original data with results of calculation program using Bacterial Foraging Optimization Algorithm method shown in table above, it can be concluded that to meet load demand approaching 3321 MW, the resulting cost is minimum compared to original data which is Rp. 2.927,621 per hour. This minimum cost can save Rp. 24,643 per hour.

CONCLUSIONS AND SUGGESTION

Conclusions

Based on the analysis and discussion that has been done, then obtained some conclusions as follows :

  1. Bacterial Foraging Optimization Algorithm method resulting cost of generation units (Component C) is more optimal than the cost derived from original data.
  2. Comparison of generation units combination between original data and results of the program clearly enough. Cost generated by the application is smaller than cost of the original data even though the difference is not too large.
  3. The application is very helpful for user in simplifying the calculation of optimal cost of generation untis.
  4. Use of Bacterial Foraging Optimization Algorithm method takes quite a long time to produce a given output, especially if a very small tolerance.

Suggestion

Some suggestions can be put forward for further research and development of application is as follows:

  1. For further research, it is suggested that researchers can further develop Bacterial Foraging Optimization Algorithm method, especially regarding the certainty of initialization parameters and the process to get the results that take a long time due to very much looping in the application.
  2. For further research, researchers are expected to include losses of transmission load or reach a wider area coverage and type of generation units are more diverse.
  3. For further research, it is to be expected that loading the results that take a long time can be resolved.

REFERENCES

Adrianti. (2010). Penjadwalan Ekonomis Pembangkit Thermal dengan Memperhitungkan Rugi Rugi Saluran Transmisi Menggunakan Metode Algoritma Genetik. Tesis tidak diterbitkan. Padang: Jurusan Teknik Elektro, Fakultas Teknik, Universitas Andalas.

Baijal, A., Chauhan, V. S., & Jayabarathi, T. (2011). Application of PSO, Artificial Bee Colony and Bacterial Foraging Optimization algorithms to economic load dispatch: An analysis. IJCSI International Journal of Computer Science Issues, 8(1), 467-470.

Brownlee, J. (2011). Clever Algorithms Nature-Inspired Programming Recipes. New South Wales: Creative Commons.

Farajianpour, S., Mohammadi, A., Travakoli, S., & Barakati, S. M. (2012). Improved Bacterial Foraging Algorithm for Optimum Economic Emission Dispatch with Wind Power. TELKOMNIKA, 10(4), 675-682.

Fauzi, M. R. (2011). Optimisasi Economic Dispatch Menggunakan Fuzzy-Bacterial Foraging Algorithm. Tesis tidak diterbitkan. Surabaya: Program Studi Magister Bidang Keahlian Teknik Sistem Tenaga, Jurusan Teknik Elektro, Institut Teknologi Sepuluh November.

Gazi, V., & Passino, K. M. (2011). Swarm Stability and Optimization. Chennai: Springer.

Indonesia Power. Diperoleh (3 Juli 2014) dari .

Jayabarathi, T., Bahl, P., Ohri, H., Yazdani, A., & Ramesh, V. (2012). A Hybrid BFA-PSO Algorithm for Economic Dispatch with Valve-point Effects. Front. Energy, 6(2), 155-163.

Joko, P., Montario, C. B., & Zamrudi. (2010). Transmission of ElectricityEnergy. Makalah Tekhnik Tenaga Listrik, 1-3.

Mahapatra, G., & Banerjee, S.(2013). A Study of Bacterial Foraging Optimization Algorithm and its Applications to Solve Simultaneous Equations. International Journal of Computer Applications (0975 – 8887), 72(5), 1-7.

Nadjamuddin, H. (2011). Bahan Ajar Perancangan Pembangkitan Tenaga Listrik. Makassar: Universitas Hasanuddin.

Nilawati, A. R. (2009). Definisi dan Simbol Flowchart. Diperoleh (29 Juni 2014) dari sumber 2 definisi dan simbol Flowchart.pdf.

Penangsang, O. (2011). Pengoperasian Optimum Sistem Tenaga Listrik. Diperoleh (20 Mei 2014) dari .

Pressman, R. S. (2010). Software Engineering: A Practitioners Approach (7th ed.). New York: McGraw-Hill Science.

Saadat, H. (1999). Power System Analysis. New York: McGraw-Hill Companies.

Shneiderman, B., & Plaisant, C. (2010). Designing The User Interface (6th ed.). Boston: Pearson.

Thomas, R. M. (2013). Survey of Bacterial Foraging Optimization Algorithm. International Jurnal of Science and Modern Engineering (IJISME), 1, 11-12.

Vijay, R. (2012). Intelligent Bacterial Foraging Optimization Technique to Economic Load Dispatch Problem. International Journal of Soft Computing and Engineering (IJSCE), 2, 55-59.

Welling, L., & Thomson, L. (2009). PHP and MySQL Web Development (4th ed.). Boston: Addison-Wesley.

Whitten, J. L., & Bentley, L. D. (2007). System Analysis and Design for The Global Enterprise (7th ed.). California: Irwin/McGraw-Hill.

Wikarsa, M. T. (2010). Studi Analisis Program Percepatan 10.000 MW Tahap 1 pada Operasi Sistem Tenaga Listrik Jawa Bali. Tesis tidak diterbitkan. Depok: Jurusan Teknik Elektro, Fakultas Teknik, Universitas Indonesia.

Wood, A. J., & Wollenberg, B. F. (1996). Power Generation, Operation and Control (2nd ed.). New York: A Willey-Interscience Publication.

Wu, G. (2012). Economic Dispatch of Hydro Power System Based on Bacterial Foraging Optimization Algorithm. International Conference of Mechanical Engineering and Automation Advances in Biomedical Engineering, 10, 189-194.

AUTHOR’S BIOGRAPHY

Melisa Astriyanti was born in 25 Mei 1991 at Jakarta.Author finished her Undergraduate Education at Universitas Bina Nusantara majoring in Information Technology and Mathematics in 2014.