基于粒子群算法的遗传算法研究  

Research on Genetic Algorithm Based on Particle Swarm Algorithm

作　　者：王文义[1] 秦广军[2] 王若雨[3] 

机构地区：[1]中原工学院计算机系,郑州450007 [2]郑州大学信息工程学院,郑州450052 [3]河南电力职工大学网络信息中心,郑州450051

出　　处：《计算机科学》2007年第8期145-147,共3页Computer Science

基　　金：国家自然科学基金"基于自适应搜索的快速运动(编号:60075006)

摘　　要：针对传统遗传算法存在的早熟收敛和易陷入局部最优解的问题,提出了一种基于粒子群算法的遗传算法,其原理是用粒子群算法来构造变异算子和进行种群分割。通过对三个典型多峰值函数的优化来评估算法性能。实验结果表明,该算法能很好地保持种群的多样性和克服早熟现象,显著提高遗传算法的收敛速度。Premature convergence and weak local optimization are two key problems existing in the conventional genetic algorithm. To overcome the shortcomings, this paper proposes an improved genetic algorithm based on the particle swarm algorithm. The basic principle is that a new mutation operator is constructed and population is divided into parts. Three typical multimodal values functions are optimized and evaluate the efficiency of the algorithm. The experimental results show, the improved genetic algorithm can not only maintain effectively the polymorphism in the colony and avoid premature, but also greatly improve the convergent speed.

关 键 词：遗传算法 粒子群算法 变异算子 种群多样性 早熟收敛 

分 类 号：O242.23[理学—计算数学]