摘要
检测器集是免疫阴性选择算法的关键所在,它直接影响到系统的效率和准确度,因此,如何确定一个最有效检测器集合是提高阴性选择算法性能的关键步骤。利用模糊思想,在定义了模糊相似度与背离度的基础上,提出了一种生成最有效检测器集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;算法确定了一个最有效的检测器集合,有效去除了检测器集中的冗余现象。仿真结果表明,该算法生成的检测器集检测范围较大,空间覆盖率高,黑洞数量大幅下降,算法具有较强的鲁棒性。
Detector set is where the shoe pinches in immune negative selection algorithm, it affects efficiency and veracity of system directly, so how to confirm the most effective detector set is a key step to improve negative selection algorithm capability. Fuzzy idea was used to bring forward an adjustable threshold immune negative selection algorithm of creating the most effective detector set based on defining fuzzy similarity and deviation. The matching threshold is adjustable in this algorithm and the number of holes can be reduced clearly through adjusting matching threshold, the algorithm confirms a most effective detector set and eliminates redundancy phenomenon in detector set effectively. The simulation results show that the detector set generated by the algorithm can detect bigger range, the rate of space coverage is high and the number of holes declines clearly, and the algorithm has better robustness.
出处
《系统仿真学报》
CAS
CSCD
北大核心
2008年第17期4596-4600,共5页
Journal of System Simulation
基金
国家自然科学基金资助项目(60305007)
关键词
阴性选择
模糊相似度
检测器集
黑洞
negative selection
fuzzy similarity
detector set
holes
