基于小波消噪变异的浮点数编码遗传算法  

Float Code Genetic Algorithm Based on Wavelet Denoising Mutation

作　　者：崔明义[1] 

机构地区：[1]河南财经学院信息学院,郑州450002

出　　处：《计算机工程》2010年第2期192-193,213,共3页Computer Engineering

基　　金：河南省基础与前沿技术研究计划基金资助项目(082300410100)

摘　　要：在遗传算法诸多编码中,浮点数编码具有其他编码所不具备的优势。针对浮点数编码在遗传操作中所产生的噪声和其对算法性能的影响,考虑变异操作在遗传算法中的重要作用,提出基于小波消噪变异的浮点数编码遗传算法,从理论上证明小波对浮点数编码噪声的分解,用小波实现其消噪变异。实验结果表明,该方法理论上是可靠的,方法上是可行的。In many Genetic Algorithm（GA） codes,Float Code （FC） is super to other codes in use. This paper is about which noise and its influence on GA performance are ignored by researches in genetic operation. It takes into account which mutation plays an important role of improving GA performance. Float Code Genetic Algorithm based on Wavelet denoising Mutation（FCGAWM） is proposed. Decomposing of FC noise is proved with wavelet in theory. FC denoising mutation is implemented in it. Experimental results indicate that the method is credible in theory and is feasible in technique.

关 键 词：小波 浮点数编码 遗传算法 消噪变异 

分 类 号：TP301.6[自动化与计算机技术—计算机系统结构]