一种基于中间结构层的分层粒子群算法

Hierarchic particle swarm algorithm based on layer of intermediate structure

针对标准粒子群算法易陷入局部极值和优化精度较低的缺点,结合复杂系统理论提出一种多层次粒子群算法,通过在算法结构中引入中间结构层,分别定义了进行大范围的较优值搜索的粒子和在较优值周围进行精细搜索的粒子,增加了粒子群的多样性,有效协调了粒子的寻优能力。采用了两种标准测试函数对算法性能进行了实验,结果表明,该算法可有效避免陷入局部最优,并在保证运行速度的同时提高了求解精度。

A hierarchic particle swarm algorithm（PSO） is proposed in order to overcome the weak ability of local search of standard PSO algo- rithm. The proposed algorithm based on layer of intermediate structure and the particles is divided into two types. One type is used to search optimum solution fast, the other type is used to search optimum solution carefully. This method can increase diversity of particles and reduce the possibility of local minimum. The experiment performance of the propose algorithm is compared with the performance of the standard PSO using two benchmark functions. The experimental simulation results indicates that the hierarchic particle swarm algorithm can escape from local opti- mal solutions efficiently and the global search ability is better than standard PSO.