改进的快速遗传算法在函数优化中的应用  

Application of improved fast-convergent genetic algorithm in function optimization

作　　者：周勇 胡中功[2] ZHOU Yong;HU Zhonggong(School of Information Engineering,Yangtze University College of Technology&Engineering,Jingzhou 434020,China;School of Electrical and Information Engineering,Wuhan Institute of Technology,Wuhan 430205,China)

机构地区：[1]长江大学工程技术学院信息工程学院,湖北荆州434020 [2]武汉工程大学电气信息学院,湖北武汉430205

出　　处：《现代电子技术》2018年第17期153-157,共5页Modern Electronics Technique

基　　金：云南省教育厅科学研究基金(2016YJS020);长江大学工程技术学院科研基金项目(2018ky08)~~

摘　　要：遗传算法作为一种模仿生物自然进化过程的全局随机优化算法,在工程中已得到广泛应用。但普通遗传算法易存在早熟及收敛速度慢等缺点。提出一种快速收敛的改进遗传算法,该算法从全局出发,对初始群体生成、遗传选择、交叉和变异算子操作等几个方面做出改进,其中重点对交叉率和变异率进行优化,实现交叉率和变异率按个体适应度以S曲线和高斯分布曲线形式进行非线性自适应调整。通过案例仿真分析,证明了该方法的可行性和有效性,且具有更快的收敛速度和更可靠的稳定性。Genetic algorithm(GA),as a global stochastic optimization algorithm to simulate the natural evolution process of biology,has been widely used in engineering field,but the common GA has slow convergence speed and is easy to fall into premature convergence.Therefore,an improved fast-convergence GA is proposed to improve the items of initial population generation,genetic selection,operators crossover and operators mutation proceeding from the overall situation.The crossover rate and mutation rate are optimized emphatically to realize the nonlinear adaptive adjustment in the forms of the S curve and Gaussian distribution curve according to individual fitness.The analysis results of case simulation show that the method is feasible and effective,and has fast convergence speed and high stability.

关 键 词：遗传算法 高斯分布 自适应 收敛 性能仿真 函数优化 

分 类 号：TN911.1-34[电子电信—通信与信息系统] TP18[电子电信—信息与通信工程]