一个SPEA改进算法及其收敛性分析

An Improved SPEA and its Convergence

SPEA是一种多目标优化算法。与其它多目标进化算法相比,SPEA算法具有设置参数少、解在空间分布均匀等优点。本文引入多点交叉和Cauchy变异对SPEA算法的收敛速度进行了改进,并对其收敛性进行了分析,文中给出的仿真算例证实了改进方法的有效性。

SPEA (Strength Pareto-Optimal Evolutionary Algorithm) is a Multi-Objective Evolutionary Algorithm (MOEA). Compared with other MOEA's, it has the advantages of fewer parameters and well-distributed solutions in objective function spaces. In this paper, a new multiple-chromosome crossover operator and a Cauchy mutation opera- tor are introduced into SPEA to speed up the algorithm. The convergence of the algorithm is also presented for the first time. The effectiveness of the improved SPEA is proved by our experiments.