【作者】 王光辉； 陈杰； 潘峰；

【Author】 Wang Guang-Hui,Chen Jie,Pan Feng Department of Automatic Control,School of Information Science and Technology,Beijing Institute of Technology,Beijing 100081,P.R.China

【机构】 北京理工大学信息科学技术学院自动控制系；

【摘要】 对于动态环境问题的优化不再是在解空间中找到一个最优点,而是要尽可能的在解空间中跟踪运动变化的最优点。作为一种基于群智能的并行随机优化方法,粒子群优化算法(PSO)在优化求解问题中体现出了良好的性能。针对基本粒子群优化算法容易收敛到局部位置而失去跟踪能力,本文提出了多种群协同粒子群优化算法(CmSPSO),当两子群的最优距离小于排斥半径时,两种群协同初始化来搜索未知解空间。实验表明,该算法简单有效,能在保持种群收敛性的基础上提高粒子多样性,能较好地跟踪变化的最优位置。更多还原

【Abstract】 Optimization problem in the dynamic environment is not only to locate a optimum,but track the moving optimum as close as possible.Particle Swarm Optimizer,a kind parallel random optimization method based on swarm intelligence,exhibits good performance for optimization problem.Particles converge at the local place quickly in basic Particle Swarm Optimizer(PSO) ,led to lose the ability to tracking the dynamic optima.To solve the problem,it’s proposed the Cooperative Multi-Swarms Particle Swarm Optimizer(CmSPSO) ,which can cooperatively reinitialize a portion of particles in each swarm,which the distance between them is less than exclusive radius.Experiments have been made to demonstrate the simplicity and effectiveness,maintained the convergence and enhanced the diversity,and obtained the good effect at tracking the dynamic optima.更多还原