基于高频变异免疫进化算法的多峰值函数优化  

Multi-modal Function Optimization Based on Immune Evolutionary Algorithm with High Frequency Variation

作　　者：岳永恒[1] 赵强[1] 

机构地区：[1]东北林业大学,哈尔滨150030

出　　处：《森林工程》2010年第2期35-38,共4页Forest Engineering

基　　金：黑龙江省自然科学基金项目(E200605)

摘　　要：多峰值函数优化问题一直是数值优化领域中研究的难点。本文提出一种基于高频变异的改进免疫进化算法,通过克隆选择高频变异抗体来增加群体多样性,又赋予抗体生命期抑制群体暴增,此外引入变异因子提高抗体识别抗原的速度。该算法即克服克隆选择算法的收敛速度慢,又克服遗传算法易陷入局部最优解的不足。对三个多峰值函数优化仿真,并和标准遗传算法、改进遗传算法和佳点遗传算法的计算结果进行比较,证明该方法可以快速收敛到全局最优解。The problem of multi - modal function optimization has been a difficulty in the field of numerical optimization. An improved immune evolutionary algorithm based on high frequency variation was proposed that the elonal expansion was applied to increase population diversity, the life span of antibody was given to suppress population explosion and the mutation factor was introduced to improve the speed of antibodies recognizing antigens. This algorithm could overcome the shortages of other algorithms that the clonal selection algorithm converged slowly and the genetic algorithm tended to fall into local optimal solution. Three multi - modal functions were optimized and simulated, then the results of this algorithm were compared with that of standard genetic algorithm, improved genetic algorithm and good point genetic algorithm, which indicated that this algorithm could quickly converge and get the global optimal solution.

关 键 词：克隆选择 高频变异 变异因子 生命期 

分 类 号：TP312[自动化与计算机技术—计算机软件与理论]