A Particle Swarm Optimizer with Chaos Dynamics for the global Optimization of Multimodal Function
Inspired by the classical particle swarm optimization (PSO) method and chaotic theories, this work presents a novel chaotic PSO (CPSO). Through adding Iterative Chaotic Map with Infinite Collapses (ICMIC) perturbation terms to the velocity update equation, we derive a new update velocity formula for a particle. The chaotic sequence generated by ICMIC can be expected to diversify the PSO population and prevent premature to local minima when a particle searches around the combinational solution of pbest and gbest. The chaotic search is firstly to implement for every one particle. Then, the chaotic dynamics is decaying over generations. Finally, the optimization process will be dominated for global optimization by the gradient descent algorithm. The novel chaotic PSO integrate the new chaotic dynamics and gradient descend for global optimization. Experiments are conducted on multimodal test function. The results demonstrate good performance of the CPSO in solving the global optimization of multimodal problems when compared with other two variants of the PSO.
chaotic particle swarm optimization (CPSO) chaos Lyapunov exponent correlation dimension
Yuyao he Huidang Zhang
College of Marine, Northwestern Polytechnical University, Xi an, China 710072
国际会议
第四届亚太地区混沌控制与同步会议(The Fourth Asia-Pacific Workshop on Chaos Control and Synchronization)
哈尔滨
英文
160-165
2007-08-24（万方平台首次上网日期，不代表论文的发表时间）