Author(s)

F.S. Smith & Q. Shen

Abstract

This paper outlines a basic diagnostic guidance mechanism that uses the experience of previous fault diagnoses to guide a fault diagnostic system in trying to identify the current faults. This is achieved by using a modied self-organising fuzzy logic controller to abstract the previous faulty behaviours in order to provide this guidance. The resultant technique is applied to nding faults in a chemical engin- eering system. 1 Introduction There are several existing methods for diagnosing faults in con- tinuous systems. One approach is Mimic (Dvorak1) which at- tempts to nd faults by determining which components could have caused the faulty behaviour and then simultaneously track each possible fault model with the aim of exonerating as many candidates as possible. The c

Keywords