一种并行的自适应量子粒子群算法

Method of Parallel Adaptive Quantum Particle Swarm Optimization

针对粒子群算法存在易陷入局部最优解的问题,提出了一种并行的自适应量子粒子群算法。通过共享粒子的两个极值,将改进后的自适应粒子群算法和边界变异的量子粒子群算法并行搜索,有效地克服了标准粒子群算法的缺陷。测试结果表明,该算法在精度和全局最优解的找寻速度方面有了很大的提高。

A new parallel adaptive quantum particle swarm opitimzation algorithm is proposed in this paper to solve the problem that standard particle swarm optimization（PSO） algorithm may easily trap into local optimal points and may obtain exact solutions at the late of the iteration with difficultly. By sharing the two extreme values of the particles, the proposed method is able to adaptively search their optimum solutions in parallel by combination of an improved adaptive PSO with a quantum Particle Swarm Optimization of boundary variation. It is proved effectively to overcome the shortcomings of standard PSO. Test results show that the accuracy and the velocity of global search for optimal solutions have been greatly improved.