一种基于Alopex和分布估计算法融合的进化算法及其在参数估计中的应用

摘要

A new evolutionary optimization algorithm (EDA-Alopex) based on Alopex and estimation of distribution algorithm (EDA) was constructed. In EDA-Alopex, EDA was embedded into Alopex-based evolutionary algorithm (AEA) for the purpose of improving the performance of AEA, owing to a fast convergence and unique evolutionary model owned by EDA. The proposed algorithm integrated the correlation information between different individuals obtained by AEA algorithm with the global probability information extracted by EDA, which can guide the population to evolve toward more promising domain. The results show that the search performance of EDA-Alopex improved by EDA gains a relatively great improvement, particularly its convergence speed has obvious advantages compared with the single AEA algorithm.%构造了一种基于Alopex(Algorithm of pattern extraction)和分布估计算法(Estimation of distribution algorithm,EDA)相融合的进化算法EDA-Alopex.该算法将分布估计算法嵌入到一种基于Alopex的群智能进化算法(Alopex-based evolutionary algorithm,AEA)中,利用分布估计算法收敛速度快及与传统进化算法进化模式不同的特点来改进AEA算法.新算法综合了AEA算法搜索得到的个体间相关性信息和EDA搜索过程中得到的全局概率信息,能够更好地指导种群向有利的区域进化.仿真结果表明:EDA改进的EDA-Alopex算法搜索性能与AEA算法的搜索性能相比有较大提高,特别是其收敛速度与AEA算法相比有明显提高.