外部种群完全反馈的元胞差分算法设计及应用
发布时间：2018-10-24 23:01
【摘要】：针对传统进化算法在求解多目标优化问题时存在多样性和收敛性不佳的问题,提出一种外部种群完全反馈的元胞差分算法。对标准元胞差分算法进行改进,在每一代进化之后,根据秩与k最近邻距离对外部种群进行修剪,并将修剪后的整个外部种群随机分配到二维网状结构,在原有变异操作中引入新的扰动来避免算法陷入局部最优。通过对6个基准函数进行测试表明,新算法相对于其他3种典型算法具有更好的前端覆盖性,新的变异方式能提高算法跳出局部最优解的能力。通过工程实例验证了所提算法的可行性与有效性。
[Abstract]:Aiming at the diversity and poor convergence of traditional evolutionary algorithm in solving multi-objective optimization problems, a cellular difference algorithm with complete feedback of external population is proposed. The standard cellular difference algorithm is improved. After each generation evolves, the external population is pruned according to the distance between rank and k nearest neighbor, and the whole external population is randomly assigned to the two-dimensional mesh structure. A new disturbance is introduced into the original mutation operation to avoid the algorithm falling into local optimum. The test of six benchmark functions shows that the new algorithm has better front-end coverage than the other three typical algorithms, and the new mutation method can improve the ability of the algorithm to jump out of the local optimal solution. The feasibility and effectiveness of the proposed algorithm are verified by an engineering example.
【作者单位】： 浙江工业大学机械工程学院;
【基金】：浙江省自然科学基金资助项目(LY16G010013) 国家自然科学基金资助项目(71371170,71301148) 国家863计划资助项目(2015AA043002)~~
【分类号】：TP18
本文编号：2292830
[Abstract]:Aiming at the diversity and poor convergence of traditional evolutionary algorithm in solving multi-objective optimization problems, a cellular difference algorithm with complete feedback of external population is proposed. The standard cellular difference algorithm is improved. After each generation evolves, the external population is pruned according to the distance between rank and k nearest neighbor, and the whole external population is randomly assigned to the two-dimensional mesh structure. A new disturbance is introduced into the original mutation operation to avoid the algorithm falling into local optimum. The test of six benchmark functions shows that the new algorithm has better front-end coverage than the other three typical algorithms, and the new mutation method can improve the ability of the algorithm to jump out of the local optimal solution. The feasibility and effectiveness of the proposed algorithm are verified by an engineering example.
【作者单位】： 浙江工业大学机械工程学院;
【基金】：浙江省自然科学基金资助项目(LY16G010013) 国家自然科学基金资助项目(71371170,71301148) 国家863计划资助项目(2015AA043002)~~
【分类号】：TP18
【相似文献】
相关期刊论文 前3条
1 肖刚;谢红;;基于二进制差分算法的HWSN目标覆盖研究[J];应用科技;2013年06期
2 彭奇;李伟华;;牺牲层腐蚀改进模型与模拟研究[J];传感技术学报;2008年02期
3 徐郡,储方杰;基于BP网时间差分算法的设计与应用[J];系统仿真学报;2003年10期
相关硕士学位论文 前1条
1 张天威;进化差分算法在联合补货模型中的应用[D];哈尔滨理工大学;2017年
，本文编号：2292830
本文链接：https://www.wllwen.com/kejilunwen/zidonghuakongzhilunwen/2292830.html
