應用免疫演算法於發電機組維修排程之研究

Hydro-Thermal Generator Maintenance Scheduling via Immune Algorithm

林融玨(Jung-Chueh Lin)

指導教授 ： 蕭瑛東

淡江大學/工學院/電機工程學系碩士班/碩士(2007年)

https://doi.org/10.6846/TKU.2007.00610

摘要

這幾年來台灣地區經濟快速成長，使得用電量急遽增加。同時，加上能源短缺及電廠的興建受到環保意識抬頭受阻，致使備轉容量明顯偏低。因此,發電機組的維修排程計畫就在電力系統運轉規劃上扮演一個很重要的角色。發電機維修排程主要目的乃在於能獲得最大備轉容量與最低發電成本情況下，求得各發電機組之維修順序與時間。

本論文將擬定維修排程計畫，並以均化備轉率為目標函數，考量實際系統的限制條件，如檢修間隔、人力限制、電力平衡等。雖然在以往的研究中，已有許多方法使用在此問題上，例如動態規劃法、整數規劃法及支界法等，但是由於這些方法所需的求解時間會隨機組數的增加呈指數關係成長，因此都比較適合於小規模系統上使用。

在本論文中，我們提出免疫演算法(Immune Algorithm)，來求解機組的維修排程問題。免疫演算法是應用抗體及抗原在免疫系統運作模式來求解最佳化問題，其中，抗體及抗原相當於最佳化問題中求解空間的一解和目標函數。利用抗體族群相似程度之關係，避免陷入局部最優解的可能性，使得在求解空間的搜尋過程中，能夠快速收斂且找到全域最佳解，結果顯示，免疫演算法對於機組的維修排程問題而言，應不失為一個很好的分析工具。

關鍵字

免疫演算法 ； 維修排程

並列摘要

Recently, the reason that economics in Taiwan grow quickly makes load demands increase rapidly. In addition, the energy is hard up and the rise of environmental protection makes the difficulties of generating unit system, the spinning reserve is scant obviously. Therefore, the maintenance scheduling plays an important role within the planning of power system operation.

In this thesis, we plan the maintenance scheduling and levelize the spinning reserve rate to be the objective function. We also consider realistic constraints, such as maintenance alternate interval, crew constraints, power balance requirement, etc. Although departed methods such as dynamic programming, integer programming, and branch bound method can solve small scale problems, the rise in execution time of these methods is exponential with the number of generating units.

In this thesis, we render Immune Algorithm to solve the maintenance scheduling problem. Immune Algorithm is used to solve the optimal problem via the operation of the antibody and the antigen in immune system. The antibody is taken as the solution of the optimal problem and the antigen is taken as the objective function of the optimal problem. To prevent the local optimal solution, we can find out the global optimal solution rapidly during searching for solution space by the diversity of antibody populations. The result obtained from analysis proves that Immune Algorithm is a good method for maintenance scheduling problem.

並列關鍵字

Immune Algorithm ； Maintenance Scheduling

參考文獻

[6] H. H. Zurn and V. H. Quintana, “Generator Maintenance Scheduling via Successive Approximation Dynamic Programming,” IEEE Transactions on Power Apparatus and Systems, Vol. PAS-94, No. 2, pp. 666-671, March/April( 1975).

[7] Z. Yamayee, “A Computationally Efficient Optimal Maintenance Scheduling Method,” IEEE Transactions on Power Apparatus and Systems, Vol. PAS-102, No. 2, pp. 330-338, February(1983).

[8] G. T. Egan, T. S. Dillon, K. Morsztyn, “The Deterministic of Optimal Maintenance Schedules in Power Systems Using the Branch and Bound Method of Integer Programming,” CIGRE 32-75-GT01-2 (1975).

[9] T. Satoh and K. Nara, “Maintenance Scheduling by Using Simulated Annealing Method,” IEEE Transaction on Power Systems, Vol. 6, No. 2, pp. 850-856, May (1991).

[14] Ge Hong, Mao Zong-Yuan, “Immune Algorithm,” Proceedings of the 4th World Congress on Intelligent Control and Automation, June 10-14, 2002, Shanghai, P.R.China.