一种改进人工免疫的飞行状态规则提取方法  

Method of rule acquirement of flight state based on improved AIS

作　　者：苏晨[1] 倪世宏[1] 王彦鸿[2] 

机构地区：[1]空军工程大学工程学院,西安710038 [2]西北工业大学自动化学院,西安710072

出　　处：《计算机工程与应用》2011年第3期237-239,共3页Computer Engineering and Applications

摘　　要：针对基本人工免疫系统寻优性能的不足,提出映射突变机制,将抗体信息映射为多维空间中的坐标点,然后依据各抗体相应的亲和力和相对位置对其进行移动,从而扩大了搜索范围,提供了更为高效的抗体产生方法。通过测试函数验证,映射突变机制的引入提高了寻优的精度和收敛速度。将改进的人工免疫系统用于飞行数据规则提取,仿真实验结果表明,基于改进人工免疫系统所提取的规则简洁、正确、可靠。Aimed at the shortcoming of the Artificial Immune System（AIS）,a mapping-mutation mechanism is put forward.The information in the antibodies is mapped to the location points in the multi-dimension space,then each point is moved based on its affinity and relative position,thus the search range is expanded and a more efficient method for generating antibodies is provided.The experiment using Rosenbrock function shows that both the precision and the convergence speed are better.Then this improved AIS is used to acquire rule from flight data.The experimental results show that the rule acquired based on improved AIS is simple,accurate and trusty.

关 键 词：人工免疫系统 映射突变 规则提取 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]