多群体遗传算法动态调整区间的参数优化方法  

PARAMETER OPTIMIZATION FOR DYNAMIC ADJUSTMENT INTERVAL OF MULTIPLE GENETIC ALGORITHM

作　　者：关晓颖[1] 谢盛嘉[2] 陈果[3] Guan Xiaoying;Xie Shengjia;Chen Guo(School of Software,Guangdong Food and Drug Vocational College,Guangzhou 510520,Guangdong,China;School of Applied Art Design,Guangdong Women s Polytechnic College,Guangzhou 511450,Guangdong,China;College of Civil Aviation,Nanjing University of Aeronautics and Astronautics,Nanjing 210016,Jiangsu,China)

机构地区：[1]广东食品药品职业学院软件学院,广东广州510520 [2]广东女子职业技术学院应用设计学院,广东广州511450 [3]南京航空航天大学民航学院,江苏南京210016

出　　处：《计算机应用与软件》2022年第5期273-282,312,共11页Computer Applications and Software

基　　金：国家自然科学基金项目(51675263);国家科技重大专项(2017-IV-0008-0045);广东食品药品职业学院自然科学类课题(2016YZ008)。

摘　　要：针对参数初始区间不能准确定义的参数优化问题,提出一种多群体遗传算法动态调整区间的参数优化方法。该方法采用多群体遗传算法,设计差分变异和均匀变异相结合的变异算子,进行区间外探索和区间内开发,并设计参数区间动态调整的策略,从而实现降低对参数初始取值区间准确性的要求,同时大大提高了算法的求解精度和求解速度。利用标准函数集验证了该算法的正确及有效性。Aiming at the problem that the initial interval of parameters cannot be accurately defined,this paper proposes a parameter optimization for dynamic adjustment interval of multiple genetic algorithm.The method utilized the multiple genetic algorithm,and designed a mutation operator combining the differential mutation and uniform mutation for out-of-interval exploration and in-interval development.We designed the strategy of dynamic adjustment for parameter intervals,which reduced the dependence on the accuracy of the initial parameter interval,and greatly improved the accuracy and speed of the algorithm.The correctness and effectiveness of the proposed algorithm were verified by using a standard function set.

关 键 词：多种群 遗传算法 参数区间 动态调整 参数优化 

分 类 号：TP27[自动化与计算机技术—检测技术与自动化装置]