动态环境下一种改进的小生境粒子群算法

Improved niche Particle Swarm Optimizer in dynamic environment

为了提高在动态环境下追踪变化的极点的可靠性和精确性的能力,避免算法收敛于一个最优解,提出了一种改进的小生境微粒群算法。使用DF1(Dynamic Function1)生成的复杂动态环境对这种算法进行了验证,并与经典的APSO(Adaptive Particle Swarm Optimizer)算法进行了对比,实验结果表明了该算法的有效性。

The purpose of this paper is to present a modified Panicle Swarm Optimization（PSO ） algorithm applied to the complex dynamic environment.The method presented is defined as Improve Niche Panicle Swarm Optimizer（ IN-PSO ）.It can improve the reliability and accuracy while tracking dynamic pole in dynamic environment and avoid converge to a optimality.The environment used in these experiments is generated by Dynamic Function #1（DF1）.The results of the experiments elucidate that INPSO is more adaptive than Adaptive Panicle Swarm Optimizer（ APSO ）.