基于元胞自动机改进的微粒群算法

An improved Particle swarm optimization based on Cellular automata

该文在分析微粒群算法局部最好模型几种邻域结构特点的基础上,提出了基于元胞自动机改进的微粒群算法。该算法从元胞自动机的建模思想出发,指出了微粒群算法本身就是一个元胞自动机,从而利用元胞自动机的理论对微粒群算法进行分析改进。实验结果表明,该算法不仅在单峰函数和多峰函数的优化中表现出了较好的性能,而且还适合比较广泛范围函数的优化。

Based on the analysis of various population topologies＇ characteristic of the Lbest population on particle swarm optimization （PSO）, this paper proposes an improved particle swarm optimization based on Cellular Automata. Associating the modeling idea of cellular automata, the algorithm points out that PSO algorithm itself is a cellular automaton, as a result, the algorithm uses theory of cellular automata to analyses improvement of PSO algorithm. Experimental simulations show that the presented algorithm not only shows a better performance in both unimodal functions and muhimodal functions, but also results in best performance on a range of functions.