基于免疫算法的移动机器人路径规划研究  

Research of Mobile Robot Path Planning Based on Immune Algorithm

作　　者：陆州[1] 李琳[1] 邹炎飚[1] 

机构地区：[1]华南理工大学机械与汽车工程学院,广东广州510640

出　　处：《机床与液压》2012年第11期25-28,共4页Machine Tool & Hydraulics

基　　金：国家863重点课题资助项目(2009AA043901-3);粤港关键领域联合招标项目(20090101-1);广州白云区科技攻关项目资助(2009-S2-46)

摘　　要：针对障碍物已知的全局路径规划问题,采用几何坐标法建立机器人工作的环境模型,提出了一种包含否定选择、克隆选择、接种疫苗等免疫思想的路径搜索方法,详细介绍了该算法的主要设计思想并给出了算法的具体流程。在亲和力构建时引入抗体浓度的概念,保持群体在进化过程中的多样性;提出了一种新的接种疫苗方法,大大改善了抗体的品质。与基于传统遗传算法的路径规划方法进行对比仿真实验,结果表明免疫算法有效地防止局部收敛,能更快地收敛到全局最优解。Aiming at the problem of global path planning under the condition of known obstacle, the environment model for mo- bile robot was built by using the coordinate geometry method. Some immune thoughts for path planning were proposed which included negative selection, clonal selection, and vaccination and so on. The main design ideas and specific steps of this algorithm in detail were also introduced. The concept of antibody concentration was introduced when affinity constructing, which in evolution process kept the diversity of the group. And a new vaccination method was researched to improve the quality of the antibody. The results manifest that by contrasting the simulation experiment between traditional genetic algorithm and immune algorithm, the convergence can be prevent effectively by the immune algorithm and the global optimum solution gotten more quickly.

关 键 词：免疫算法 路径规划 接种疫苗 克隆选择 

分 类 号：TP242.2[自动化与计算机技术—检测技术与自动化装置]