Abstract: To the problems of slow convergence and easy to fall into local optimization appeared in standard particle swarm optimization, this paper proposes a particle compaction and scheduling based particle swarm optimization (PCS-PSO). Firstly, this paper presents the regulations of particles’ compaction and scheduling. In order to avoid particles to stay in local optimization, PCS-PSO evaluates dynamically particle’s compaction and schedules the particle when the value of the particle’s compaction is beyond the threshold. Compared with standard particle swarm optimization and other optimization algorithms using 11 benchmark functions, the experimental results show that PCS-PSO has better behaviors in convergence accuracy and speed.

Key words: particle swarm optimization, local optimization, compaction, scheduling, accuracy of convergence, speed of convergence

摘要： 针对标准粒子群优化算法存在收敛速度慢和难以跳出局部最优等问题，提出了一种基于紧凑度和调度处理的粒子群优化算法。给出了粒子紧凑度和调度处理的概念和方法，通过动态评价粒子群中各粒子间的紧凑程度，从而确定调度的粒子，进而对其进行调度处理，避免粒子陷入局部最优。对11个常见的标准函数进行测试，并与标准粒子群算法和其他改进算法进行对比，实验结果表明，基于紧凑度和调度处理的粒子群优化算法具有较高的寻优精度和较快的收敛速度。

关键词: 粒子群优化算法, 局部最优, 紧凑度, 调度处理, 寻优精度, 收敛速度