摘要
通过对人工免疫系统中阴性选择算法机理的分析,利用模糊思想,定义了模糊相似度,并在此基础上,提出了一种基于模糊思想的变阈值免疫阴性选择算法.该算法匹配阈值可变,采用调整匹配阈值的方法大幅降低黑洞数量;在满足一定模糊相似度的前提下,实现了带控制参数的模糊匹配,模糊程度可控.仿真结果表明,该算法生成的成熟检测器检测范围较大,空间覆盖率高,黑洞数量大幅下降,同时检测率有显著提高,算法具有较强的鲁棒性.
This paper analyses mechanisms needed for a negative selection algorithm in an artificial immune system and defines fuzzy similarity using fuzzy logic. Based on this, an adjustable threshold immune negative selection algorithm using fuzzy ideas is proposed. In this method, the adjustable matching threshold is used to greatly reduce the number of black holes, improving fuzzy matching for control parameters with the prescribed fuzzy similarity. The fuzzy degree can also be controlled. Simulation results show that the mature detector generated by the algorithm can detect wider ranges with high spatial coverage and fewer holes. Also, the detection rate is significantly improved and the algorithm is more robust.
出处
《哈尔滨工程大学学报》
EI
CAS
CSCD
北大核心
2007年第11期1222-1227,共6页
Journal of Harbin Engineering University
基金
国家自然科学基金资助项目(60305007)
关键词
人工免疫系统
阴性选择
模糊相似度
阈值
黑洞
artificial immune system
negative selection
fuzzy similarity
threshold
black holes
