基于种群混合迁移策略的并行量子遗传算法
Parallel quantum genetic algorithm based on multiple population-migration strategies
陆涛 1管荑 2贾鹏 1曲志坚 3王子灵3
作者信息
- 1. 南京南瑞信息通信科技有限公司通信业务部,江苏南京 211100
- 2. 国网山东省电力有限公司 电力调度控制中心通信处,山东济南 250012
- 3. 山东理工大学计算机科学与技术学院,山东淄博 255049
- 折叠
摘要
针对量子遗传算法求解大规模优化问题存在收敛速度慢、易于陷入局部最优等问题,改进量子遗传算法.设计一种种群混合迁移机制促进算法的种群多样性,采用仿TriBA种群结构、双精英种群、重生种群、自适应迁移算子、个体竞争排挤算子以及随机失活机制,提高算法的局部勘测能力和全局寻优能力.利用Spark框架实现算法在分布式集群环境下的运算.改进2-opt&-R优化算法,通过引入高斯变异提高算法的局部搜索能力,缩小算法的搜索空间.实验结果表明,改进后的算法在全局优化能力、收敛速度、运行速度和求解稳定性等方面均有大幅度提升.
Abstract
A multiple population-migration strategies mechanism was designed to promote population diversity,and the imitating TriBA population structure,dual elite population,rebirth population,self-adaptive migration operator,individual competition crowding out operator,and random dropout scheme were designed to improve the local survey and global optimization capability of the algorithm.The Spark framework was employed to realize the algorithm in a distributed cluster environment.An improved 2-opt&R optimization algorithm was also proposed.To improve the local search ability of the algorithm,Gaussian mutation was introduced to further reduce the search space.Experimental results show that the proposed algorithm greatly improves its global optimization capability,convergence speed,execution speed,and optimization stability.
关键词
量子遗传算法/种群迁移/Spark框架/并行计算/收敛速度/全局优化/搜索空间Key words
quantum genetic algorithm/population migration/Spark framework/parallel computing/convergence speed/global optimization/search space引用本文复制引用
基金项目
国家电网有限公司总部科技基金项目(5700-202116378A-0-0-00)
山东省高等学校青年创新团队发展计划基金项目(2019KJN48)
出版年
2024
NETL
NSTL
万方数据