吸收变异的粒子群算法及应用

Convergence mutation particle swarm optimization and application

摘要:

提出了一种改进的粒子群算法,称为Convergence mutation PSO(CM_PSO)。该算法根据粒子群中各粒子的适应值动态地调整粒子数量,把部分适应度差的粒子适当汇聚到适应度好的区域,在不损失全局搜索能力的同时,增强局部搜索能力。在将该算法应用于雷达组网误差配准问题中,提出了一种基于CM_PSO的误差配准技术。仿真表明,该算法能有效配准,与对比方法相比较,具有更好的收敛精度和更快的进化速度。

Abstract:

In this paper, an improved PSO named convergence mutation PSO (CM_PSO) is introduced. CM_PSO adjusts dynamically the number of particle according to the fitness of each particle. Under the request of not losing the capability of global searching, the part of particles of the worse fitness is focus into the better fitness area in order to strengthen the capability of detail searching. The simulation for radar system error shows that compared with the contrast method, CM_PSO algorithm has the better convergence accuracy and faster evolution velocity.