GENERALIZED HALF-SPLIT SEARCH FOR MODEL-BASED DIAGNOSIS

A major step in model-based fault diagnosis is the generation of candi date components which might be responsible for the observed symptom of malfunction. After the candidates are determined, each component can then be examined in turn. It is useful to be able to choose the most l ikely candidate to focus on first so that the faulty parts can be loca ted sooner. This paper presents some topological heuristics for orderi ng candidates according to their relationships with violations and cor roborations. Candidates are also reordered or eliminated when new info rmation is acquired during diagnosis. These ordering and reordering te chniques are shown to be effective for model-based diagnosis. For sing le fault cases, the average length of diagnosis is log n, where eta is the number of components. In general, the average length of diagnosis is k log n when there are k faults in the device initially.