Multiple Fault Diagnostic Strategy for Redundant System

Abstract

It is difficult to diagnose the faults, especially multiple faults, in redundant systems by traditional diagnostic strategies. So the problem of multiple fault diagnostic strategy for redundant system was researched in this paper. Firstly, the typical characters of multiple faults (minimal faults) were analyzed, and the problem was formulated. Secondly, a pair of two-tuples were applied to denote the possible and impossible diagnostic conclusion at different diagnostic stages respectively, and a multiple fault diagnostic inference engine was constructed based on Boolean logic. The inference engine can determine the system diagnostic conclusions after executing each test, and determine whether a repair action was needed, and further determine whether a next test was needed. Thirdly, a method determining the next best test was presented. Based on the proposed inference engine and test determining method, a multiple fault diagnostic strategy was constructed. Finally, a simulation case and a certain flight control system were applied to illustrate the proposed diagnostic strategy. The simulation and practical data computational results show that the presented diagnostic strategy can diagnose multiple faults in redundant systems effectively and it is of certain application value.