一、引言自六十年代初至今模拟电路故障诊断成为热门课题,出现了许多成果。模拟电路的故障字典诊断法便是其中之一。故障字典法有两个步骤,首先是建立故障字典,其次是故障分离。本文试图解决动态模拟电路的故障诊断问题。首先阐述了状态矩阵字典法的概念,用系统识别法通过测量值求解状态矩阵,从而解决了故障诊断的关键问题。在故障分离中。我们提出利用模糊决策进行,此方法明显优于 P.M.Lin 提出的方法,且物理概念明确。 I. INTRODUCTION Analog circuit fault diagnosis has become a hot topic since the early 1960s and many achievements have been made. Analog dictionary fault dictionary diagnostic method is one of them. Fault dictionary method has two steps, the first is to establish fault dictionary, followed by fault separation. This article attempts to solve the problem of dynamic analog circuit fault diagnosis. Firstly, the concept of the state matrix dictionary method is expounded. The system identification method is used to solve the state matrix through the measured values, thus solving the key problems of fault diagnosis. In the fault separation. We propose the use of fuzzy decision-making, this method is significantly better than P.M.Lin proposed method, and the physical concept is clear.