基于抗体注射的免疫遗传算法  

Immune genetic algorithm based on antibody injection

作　　者：朱思峰[1,2] 邵国金[3] 王华东[1] 

机构地区：[1]周口师范学院数学与信息科学系,河南周口466000 [2]西安电子科技大学计算机学院,西安710071 [3]河南城建学院计算机科学与工程系,河南平顶山467001

出　　处：《计算机应用》2009年第6期1536-1538,1543,共4页journal of Computer Applications

基　　金：国家自然科学基金资助项目(6057503760502043);河南省自然科学研究项目(082400440260);河南省教育厅自然科学研究项目(2008A180041)

摘　　要：标准遗传算法存在收敛速度慢、过早成熟等缺点。借鉴生物免疫系统中抗体注射免疫的理论,提出了一种基于抗体注射的免疫遗传算法(AIGA)。该算法在保留了标准遗传算法随机全局搜索能力的基础上,引进了生物免疫系统的免疫应答、抗体注射和免疫选择等机制。结合旅行商问题(TSP),给出了示范抗体的提取和注射方法,并给出了算法收敛性的理论证明。最后,用AIGA对100个城市的TSP进行了仿真计算,并将其计算过程与标准遗传算法进行了对比,结果表明该算法能有效地改善遗传算法不成熟收敛的缺陷,使收敛的速度有较大的提高。Standard genetic algorithm has some disadvantages, such as slow astringency, precocity. A new immune genetic algorithm （AIGA） based on the theory of antibody injection immune was proposed. Immunity response, antibody injection and immunity selection of biological immune systems were introduced to combine with the search ability of standard genetic algorithm in AIGA. When the method of antibody distilling and injecting used in Travelling Salesman Problem （TSP） were presented, the convergence of AIGA was proven theoretically. A simulation test of 100-city TSP was done with AIGA, and its computational process was compared with that of standard genetic algorithms. The test results show that AIGA can overcome the disadvantage of precocity and improve the speed of astringency evidently.

关 键 词：遗传算法 免疫算法 抗体 收敛性 旅行商问题 

分 类 号：TP39[自动化与计算机技术—计算机应用技术] O224.23[自动化与计算机技术—计算机科学与技术]