面向全局优化的混沌粒子群遗传算法

Chaotic Particle Swarm Optimization Based on Genetic Algorithm Oriented to Global Optimization

结合遗传算法(GA)和粒子群算法(PSO)的优点以及混沌运动的特性,提出了混沌粒子群遗传算法(CPSO-GA),并使用五个高维非线性测试函数考察此算法的性能。在固定进化代数、所调用目标函数次数接近以及固定收敛精度三种情况下对算法进行数值试验,结果表明,与其他文献中提出的算法相比,CPSO-GA能100%地找到最优解,收敛效果及寻优能力好,并能有效摆脱局部极小点,且调用目标函数次数最少,大大降低了计算量。

Combining the advantages of genetic algorithm(GA)and particle swarm optimization(PSO),and the features of chaotic motion,this paper proposes chaotic particle swarm optimization based on genetic algorithm(CPSO-GA)and uses five high-dimension nonlinear test function to check the performance of this calculation.The calculation is tested numerically under three cases:the fixed evolving algebra,the similar power of the applied objective function and the fixed convergence accuracy.The results show that,compared with the other calculations in the literature,CPSO-GA can be used to absolutely find the optimal solutions,and its convergence effect and optimization capacity are satisfying,which can effectively get rid of local minimum point,and use the objective function to the least times.In this way,the amount of calculation is greatly reduced.