对模糊免疫算法应用于模拟电路故障诊断进行了研究;首先简要介绍了免疫系统的工作机理及一些基本概念,然后在此基础上构建出一种模糊免疫算法,并将免疫算法和模糊聚类法结合起来进行故障诊断;人工免疫算法起到学习样本的作用,以寻找到各样本组的聚类中心;而模糊聚类算法则准确地完成对样本的分类任务;仿真实例表明:立足于模拟电路故障诊断字典法,该算法对模拟电路故障诊断非常有效。
A new method of fault diagnosis for analog circuits based on fuzzy--immune algorithm is proposed. At first, the operational principles of immune system and its basic conception are briefly introduced. Next an hybrid algorithm combining fuzzy clustering algorithm and immune algorithm is proposed and used for fault diagnosis of analog circuits. For the presented fault diagnosis approach, immune algorithm has been modified, its action is to find the center of training samples. Fuzzy clustering algorithm is used to classify the faults samples correctly. Finally , an example for the fault diagnosis of analog circuits using fault dictionary was shown and the simulation results indicated the validity of the proposed algorithm.