一种求解复杂多峰问题的新型粒子群优化算法研究

为提升标准粒子群算法在求解多峰复杂问题时收敛速度慢和极易陷入局部最优解等缺点, 提出一种基于球形坐标的分类学习策略粒子群算法(CLPSO-HC)。该算法给出种群运行较差粒子的确定方法, 将运行较差的粒子进行分类, 并对每类粒子给出相应的学习策略, 保证种群跳出局部最优解的能力。为减少外界扰动, 将粒子速度和位置的更新在球形坐标中进行, 提升了种群向最优解飞行的概率。对三个典型测试函数进行仿真实验, 所得结果表明CLPSO-HC相比其他几种算法有较好的收敛性。因此, CLPSO-HC可以作为求解复杂多峰问题的有效算法。

Novel particle swarm optimizer forsolving complicated multimodal problem

In order to deal with the problems of the slow convergence and easily converging to local optima, this paper proposed a classification learning PSO based on hyperspherical coordinates. It presented the method of determination of poor performance particle, and divided the swarm into three parts where introduced three learning strategies to improve the swarm to escape from local optima. Additionally, to decrease outside disturbance, it updated the particle positions and velocities in hyperspherical coordinate system, which improved the probability flying to the optimal solution. It conducted the simulation experiments of three typical functions, and the results show the effectiveness of the proposed algorithm compared with other algorithms. Consequently, CLPSO-HC can be used as an effective algorithm to solve complex multimodal problems.