论文部分内容阅读
元胞遗传算法通过限定个体之间的相互作用邻域提高算法的全局收敛率,但在一定程度降低搜索效率.文中提出一种粒子群与多种群元胞遗传混合优化算法.首先将群体分割成多个相互之间没有邻域关系的元胞子种群,适度降低算法的选择压力,从而更好地保持种群的多样性.算法的变异操作被粒子群算法替代,使得局部搜索能力明显提高.元胞群体分割和粒子群变异较好地均衡全局探索和局部寻优之间的关系.分析混合算法的选择压力和多样性变化规律.实验结果表明,该算法在保证搜索效率较高的同时还显著提高元胞遗传算法的全局收敛率且稳定性得到明显改善.
Cellular GA improves the global convergence rate by limiting the neighborhood between individuals, but reduces the search efficiency to a certain degree.In this paper, a genetic hybrid optimization algorithm is proposed for particle swarm optimization (PSO) A large number of cell populations with no neighborhood relationship between each other and moderately reduce the selection pressure of the algorithm to better maintain the diversity of the population.The algorithm of mutation operation is replaced by PSO, Population segmentation and PSO well balance the relationship between global exploration and local optimization, and analyze the selection pressure and diversity variation of hybrid algorithm.The experimental results show that this algorithm not only improves the search efficiency but also significantly improves The global convergence rate and stability of CGA are significantly improved.