Abstract: The Artificial Immune System（AIS） inspired by the natural immune system is introduced.By the analyses on the existing detector generating algorithms，we achieve the improvement of the linear time detector generating algorithm，which is widely used at present.Arrays C and C′ are constructed from two directions and are crossed to get D，which makes detectors match more strings of “nonself”.And removal of redundant detectors narrows the scale of detector set.Both mathematical analysis and experiment show that the improved algorithm reduces the size of the detector set，and the value of P_f decreases.

Key words: Artificial Immune System（AIS）, negative selection model, detector generating algorithm

摘要： 在介绍人工免疫系统基本概念的基础上，讨论了人工免疫系统中应用广泛的负选择模型。研究的重点是负选择模型中初始检测器集的生成算法，对线性检测器生成算法进行了改进。改进算法从两个方向分别构造数组C和C′，并做叉乘得到D数组，使产生的检测器能够匹配更多的“非我”字符串，去除冗余检测器缩小了检测器集的规模，包括算法的设计、性能分析和实验。理论分析与实验结果表明，改进算法降低了检测器规模，漏检概率P_f有较大下降。

关键词: 人工免疫系统, 负选择模型, 检测器生成算法