一种高效粒子群优化算法

An efficient particle swarm optimization

针对标准粒子群算法收敛速度慢和易出现早熟收敛等问题,提出一种高效粒子群优化算法.首先利用局部搜索算法的局部快速收敛性,对整个粒子群目前找到的最优位置进行局部搜索;然后,为了跳出局部最优,保持粒子的多样性,给出一个学习算子.该算法能增强算法的全局探索和局部开发能力.通过对10个标准测试函数的仿真实验并与其他算法相比较,结果表明了所提出的算法具有较快的收敛速度和很强的跳出局部最优的能力,优化性能得到显著提高.

To the problems of low searching speed and premature convergence frequently appeared in standard particle swarm optimization（PSO） algorithm, an efficient particle swarm optimization（AEPSO） is proposed in this paper. The method makes full use of the local convergent performance of the local random search algorithm to optimize the global best position of the swarm found so far. Then to go out of the local optimum in PSO and maintain the population diversity in the process of evolution, a learning operator is presented. This algorithm can enhance the exploration and exploitation ability of the algorithm. Through testing the performance of the proposed approach on a suite of 10 benchmark functions and comparing with other rneta-heuristics, the result of simulation shows that the proposed approach has better convergence rate, great capability of preventing premature convergence and superior performance.