嵌入隔离小生境技术的混沌粒子群算法

针对标准粒子群算法(standard particle swarm optimization,SPSO)无法很好平衡全局与局部搜索能力,且收敛速度较慢、易于早熟收敛等问题,提出了嵌入隔离小生境技术的混沌粒子群算法(isolation niches em-bedded in chaos particle swarm optimization,INCPSO)。利用隔离小生境技术,保证了解的多样性,同时,引入混沌搜索策略,提高了解的搜索精度和收敛速度,且避免早熟收敛。仿真试验结果表明,与标准粒子群算法和只嵌入隔离小生境技术的粒子群算法(isolation niches particle swarm optimization,INPSO)相比,嵌入隔离小生境技术的混沌粒子群算法对复杂问题的求解能力较强,寻优性能较好。

Chaos particle swarm optimization combined with isolation niche

To keep a balance of global and local searching ability and to overcome the slow convergence and prematurity of particle swarm optimization,a new algorithm named as isolation niches embedded in chaos particle swarm optimization(INCPSO) is proposed.Isolation niches technology ensures solutions diversity.Meanwhile,chaos searching which avoids premature convergence is introduced to search around the extremum.Solution accuracy and convergence speed are improved.The simulation results show that INCPSO is more effective to solve complex problems compared with the standard particle swarm optimization and isolation niches embedded in particle swarm optimization.