敏捷移動粒子群最佳化方法

Yare immigration particle swarm optimization

陳弘毅(Hong-Yi Chen)

指導教授 ： 莊堯棠

國立中央大學/資訊電機學院/電機工程學系/碩士(2012年)

摘要

敏捷遷移粒子群演算法(YIPSO)是藉由觀察於鳥群、魚群以及學生團體等群體來改善標準粒子群演算法(PSO)進而增強其收斂效率。通常一個族群中的個體常常會因為能力、興趣以及個性等等分為許多更小的團體，而這些小團體的行為通常會間接影響整個族群的表現。根據上述狀況，YIPSO選擇在標準演算法中加入兩種概念：將整個族群隨機分組成較小的族群，而每個小族群之中最好的個體將會帶領其他個體更迅速且有效率地往最佳的結果邁進。其二，再將在整個族群中表現較好的粒子當作菁英挑出，這些菁英將會比以往只有一個gbest對整個族群有著更大的影響性。在改善了標準粒子群演算法之後，考慮將其應用於一水輪機調速器之PID控制系統的參數選擇，且比較不同演算法對於其參數的選擇來讓整個系統穩定。

關鍵字

粒子群最佳化方法 ； 加速度係數 ； 隨機分組 ； 菁英

並列摘要

The yare immigration particle swarm optimization (YIPSO) is an improved method of the standard particle swarm optimization by observing behaviors of the flocks of fishes, birds and students to enhancing the performance of the swarm. There are usually a few smaller groups in the flock because of the ability, interest, individuality, etc., and these groups might affect the result of the flock. Considering thess situations, two concept are added to PSO as YIPSO. The first one is dividing the flock into smaller groups randomly, therefore the best one of each smaller group will take other individuals to the optimal way. The second part is choosing not only one best as gbest but some behaving well in the flock as elitists. Thus these individuals performing well will make a greater impact than before. After improving the original particle swarm optimization, there is a water turbine governor system with PID controller which needs for parameter choosing, so we use some different particle swarm optimizations to select the parameter of the PID controller.

並列關鍵字

sub-swarm ； acceleration coefficient ； elitist ； particle swarm optimization

參考文獻

[21] 林柏勳，胡光復，沈哲緯，辜炳寰與鄭錦桐，「最佳化方法於工程上之應用」，中興工程季刊，2009。

[1] I. Mukherjee and P. K. Ray, “A review of optimization techniques in metal cutting processes,” Computers & Industrial Engineering, Vol. 50, no.1-2, pp.15-34, 2006.

[2] M. Imiela, “High-fidelity optimization framework for helicopter rotors,” Aerospace Science and Technology, 2011.

[3] I. Averbakh, “Computing and minimizing the relative regret in combinatorial optimization with interval data,” Discrete Optimization, Vol. 2, no.4, pp.273-287, 2004.

[4] B. Srinivasan, D. Bonvin, E. Visser and S. Palanki, “Dynamic optimization of batch processes II. Role of measurements in handling uncertainty,” Computers and Chemical Engineering, Vol. 27, no.1, pp. 27-44, 2002.