利用遺傳算法搜索全局最優(yōu)的一種混合算法
A HYBRID ALGORITHM FOR FINDING GLOBAL OPTIMUM WITH GENETIC ALGORITHM
-
摘要: 該文提出了一種基于遺傳算法與梯度法相結(jié)合的混合優(yōu)化搜索方法。該算法能幫助梯度法跳出局部最優(yōu),獲得全局最優(yōu)解,并對(duì)算法的收斂性進(jìn)行了證明。最后給出的仿真結(jié)果表明了算法的有效性。
-
關(guān)鍵詞:
- 遺傳算法; 梯度法; 全局最優(yōu)解; 收斂性
Abstract: In this paper, a hybrid aJgorithm based on the combination of genetic algorithm and gradient method is proposed. It can help the gradient method to skip the local optimum and achieve the global minimum. The convergence of the hybrid algorithm is proved and the simulation result shows the effective of the algorithm at the end. -
錢(qián)富才,利用混沌搜索盒局最優(yōu)的一種方法,信息與控制,1998,27(3),232-235.[2]Z. Michalewice, C. Z. Janikow, J. B. Krawczyk, A modified genetic algorithm for optimal control problems, Computer Math. Appl., 1992, 23(12), 83-94.[3]惲為民,遺傳算法的全局收斂性和計(jì)算效率分析,控制理論與應(yīng)用,1996,13(4),455-459[4]張彤,變尺度混沌優(yōu)化方法及其應(yīng)用,控制與決策,1999,14(3),285-287.[5]R. Salomon, Evolutionary algorithm and gradient search, similarities and differences, IEEE Trans.on Evolutionary Computation, 1998, EC-2(2), 45-55. -