Computational & Technology Resources
an online resource for computational,
engineering & technology publications
Civil-Comp Proceedings
ISSN 1759-3433
CCP: 95
PROCEEDINGS OF THE SECOND INTERNATIONAL CONFERENCE ON PARALLEL, DISTRIBUTED, GRID AND CLOUD COMPUTING FOR ENGINEERING
Edited by: P. Iványi and B.H.V. Topping
Paper 49

A Cooperative Multi-Objective Island Parallel Model for Wind Farm Planning

A.L. Márquez1, C. Gil1, R. Baños1, M.G. Montoya1, F.G. Montoya2 and F. Manzano-Agugliaro2

1Department of Computer Architecture and Electronics, 2Department of Rural Engineering,
University of Almeria, Spain

Full Bibliographic Reference for this paper
A.L. Márquez, C. Gil, R. Baños, M.G. Montoya, F.G. Montoya, F. Manzano-Agugliaro, "A Cooperative Multi-Objective Island Parallel Model for Wind Farm Planning", in P. Iványi, B.H.V. Topping, (Editors), "Proceedings of the Second International Conference on Parallel, Distributed, Grid and Cloud Computing for Engineering", Civil-Comp Press, Stirlingshire, UK, Paper 49, 2011. doi:10.4203/ccp.95.49
Keywords: hybrid algorithm, threaded computation, island parallel model, multi-objective evolutionary algorithm, wind farm planning, aerogenerators.

Summary
The island parallel model [2,3], is a well known paradigm for creating parallel evolutionary algorithms. Its strengths lie in the use, in each island, of isolated environments where individuals can evolve independently of other populations. But sometimes evolutionary algorithms can get stuck in a local minima. The island parallel model solves this situation with the use of a migration operation, allowing the best subjects on each island to migrate to nearby populations, therefore spreading new genetic information into those populations and helping to avoid the local minima problem. Since most MOEAs optimize only two objectives, by using an island model it is intended to analyze different sets of objectives on different islands. To keep consistency, one objective must be common among nearby islands, while they have a different second objective. In this way the isolation and genetic diversity between islands is greatly amplified. The use of an "adaptation operation" is needed to update both the calculated objective values and the comparison operators.

These concepts are applied to a problem of wind farm planning with three objectives: maximizing energy production for the site while reducing the maximum power generation potential, and the third objective would be to reduce the standard deviation between the maximum and minimum energy production. For this purpose, the software will use statistical wind information of several meteorological stations [4] that has been acquired every ten minutes over the course of one year. With this data the software can make a prediction of the average output in a typical day, to find an optimum selection of two or three different models, between all the available aerogenerator data, to be installed in the wind farm and to optimize the different objectives of the proposed problem.

References
1
C.C. Coello, G. Lamont, D. van Veldhuizen, "Evolutionary Algorithms for Solving Multi-Objective Problems", Genetic and Evolutionary Computation, Springer, Berlin, Heidelberg, 2nd edition, 2007.
2
D. Van Veldhuizen, J. Zydallis, G. Lamont, "Considerations in Engineering Parallel MultiObjective Evolutionary Algorithms", IEEE Transactions on Evolutionary Computation, 7(2), 144-173, 2003. doi:10.1109/TEVC.2003.810751
3
E. Alba, M. Tomassini, "Parallelism and Evolutionary Algorithms", IEEE Transactions on Evolutionary Computation, 6(5), 443-462, 2002. doi:10.1109/TEVC.2002.800880
4
Q. Hernández-Escobedo, F. Manzano-Agugliaro, A. Zapata-Sierra, "The wind power of Mexico", Renewable and Sustainable Energy Reviews, 14(9), 2830-2840, 2010. doi:10.1016/j.rser.2010.07.019

purchase the full-text of this paper (price £20)

go to the previous paper
go to the next paper
return to the table of contents
return to the book description