A Fault Diagnosis Expert System Structure with Knowledge Retrieving Capability from Waveforms

Abstract

Analogue fault diagnosis can be addressed by expert systems using artificial intelligence methodologies. The structure of such an expert system and its main features are presented in this paper. As a special feature, knowledge which is obtained through the waveforms of specific parameters of the circuit under test is presented. Techniques for comparison between waveforms and conclusion extraction are described. Demonstrative examples of fault location using the expert system are given.