Abstract:

To solve the premature convergence problem of particle swarm optimization (PSO) in dealing with complex high dimensional function optimization, a novel hybrid particle swarm optimization algorithm merging crossover mutation and chaos (CMCPSO) was proposed. The main approaches included using chaos strategy to initiate positions and velocities of all particles in the design space, introducing crossover operation in each iteration to increase the diversity of particles, breaching the restrictions of local optimization points with a new chaotic disturbance mechanism and mutation operation during the later computation period. Four standard test functions were selected to have a simulation study on the proposed algorithm. The results showed that CMCPSO had a fast convergence rate and effective global optimization ability.

摘要：

针对粒子群算法在多峰函数优化中极易陷入局部最优的问题，提出一种融合交叉、变异以及混沌的新型混合粒子群算法。该算法采用混沌初始化所有粒子位置和速度，保证初始粒子在解空间均匀分布；在每代进化过程中引入交叉操作增加种群的多样性；并且在算法后期，粒子陷入局部极值时，采用一种新的自适应混沌扰动机制和变异机制，以确保粒子跳出局部最优位置。选用4个标准测试函数对所提出的算法进行对比仿真研究，结果表明，该算法具有较快的收敛速度、有效的全局寻优能力。