Abstract: Evolutionary programming has been applied with success to many numerical and combinatorial optimization problems. But algorithms with one single mutation operator often have the problem that they perform well when deal with these problems, but poor with others. In this paper, we propose an improved evolutionary programming using a mixed strategy based on Gaussion mutation and single-point mutation, which is named SPCEP. In SPCEP, each of the two mutation operators will generate one offspring. Then a compare will be conducted between them, and the best one will be chosen. Simulation results show that SPCEP is obviously superior to classical CEP and SPMEP for high-dimensional and unimodal functions.

Key words: mixed-strategy, Gaussion mutation, single-point mutation, optimization problems

摘要： 进化计算已成功地运用到各种数值优化和组合优化问题中，而运用单一变异算子的进化算法总是存在着对某种函数优化问题性能良好，对另一些却不尽人意的问题。本文提出一种基于经典进化算法和单点变异算法的混合策略进化算法SPCEP。SPCEP算法利用两种变异算子各产生一个后代个体，并选择较好的一个作为唯一的后代。实验结果表明，在处理高维单模函数时，SPCEP的性能比CEP和SPMEP有显著的提高。

关键词: 混合策略, 经典进化算法, 单点变异算法, 最优化问题