摘要
经典的粒子群是一个有效的寻找连续函数极值的方法,结合遗传算法的思想提出的混合粒子群算法来解决背包问题,经过比较测试,6种混合粒子群算法的效果都比较好,特别交叉策略A和变异策略C的混合粒子群算法是最好的且简单有效的算法,并成功地运用在投资问题中。对于目前还没有好的解法的组合优化问题,很容易地修改此算法就可解决。
The classical particle swarm optimization is a powerful method to find the minimum of a numerical function, on a continuous definition domain. The particle swarm optimization algorithm combining with the idea of the genetic algorithm is recommended to solve knapsack problem. All the 6 hybrid particle swarm optimization algorithms are proved effective. Especially the hybrid particle swarm optimization algorithm derived from across strategy A and mutation strategy C is a simple yet effective algorithm and it has been applied successfully to investment problem. It can easily be modified for any combinatorial problem for which there has been no good specialized algorithm.
出处
《中国工程科学》
2006年第11期94-98,共5页
Strategic Study of CAE
关键词
粒子群算法
背包问题
遗传算法
变异
particle swarm algorithm
knapsack problem
genetic algorithm
mutation
