Abstract: Particle Swarm Optimization（PSO） is a stochastic global optimization evolutionary algorithm. In this paper, a novel Chaos Particle Swarm Optimization algorithm（CPSO） is proposed in order to overcome the poor stability and the disadvantage of easily getting into the local optimum of the Standard Particle Swarm Optimization（SPSO）. On the one hand, the uniform particles are produced by logical self-map function so as to improve the quality of the initial solutions and enhance the stability. On the other hand, two sets of velocity and position strategies are employed, that is to say, the special velocity-position is used for the global particles, while the general velocity-position is used for the rest particles in the swarm so as to prevent the particles from plunging into the local optimum. The CPSO proposed in this paper is applied to four benchmark functions and the experimental results show that CPSO can improve the performance of searching global optimum efficiently and own higher stability.

Key words: Particle Swarm Optimization（PSO）, logical self-map, local convergence, stability, swarm intelligence

摘要： 针对粒子群优化算法稳定性较差和易陷入局部极值的缺点，提出了一种新颖的混沌粒子群优化算法。一方面，在可行域中应用逻辑自映射函数初始化生成均匀分布的粒群，提高了初始解的质量和增加了算法的稳定性；另一方面，采用两组速度-位移更新策略，即对全局最优粒子单独使用特定的速度-位移策略更新，而对其余粒子则使用常规的速度-位移进行更新，从而有效避免了算法陷入局部收敛的缺点。将该算法应用在4个基准测试函数优化中，仿真结果表明其能有效提高全局寻优的性能，且稳定性好。

关键词: 粒子群优化, 逻辑自映射, 局部收敛, 稳定性, 群体智能