并行定向扰动的混合粒子群优化算法

针对全向变异易使粒子失去已有的有利搜索信息的问题, 提出了一种并行定向变异的混合粒子群优化算法。该算法以当前群体最优位置为基准, 用变异信息矩阵和混沌位置变异矩阵对群体进行并行定向扰动, 有效利用了现有的有利搜索信息。该算法将并行定向变异与序列二次规划法融为一体, 实现了全局搜索和局部寻优的统一。仿真实验和比较分析结果表明并行定向变异混合粒子群优化算法具有良好的、稳定的优化效果。

Hybrid particle swarm optimization algorithm based onparallel directional turbulence

Aiming at the problem that omnidirectional mutant easily causes particles to lose existing beneficial searching information, the paper presented a hybrid particle swarm optimization algorithm based on parallel directional turbulence. On the basis of optimal location of the current swarm, the algorithm used mutant information matrix and chaotic position mutant matrix to exert parallel directional turbulence on the swarm, and effectively utilized existing beneficial searching information. The algorithm integrated parallel directional turbulence with sequential quadratic programming method so as to realize unification between global search and local search. The simulation and comparative results show that the hybrid particle swarm optimization algorithm based on parallel directional turbulence can achieve more excellent and stable optimization effect.