An example of fault diagnosis by means of probabilistic logic reasoning

Abstract An example of an application of qualitative model-based diagnosis is presented. In contrast to approaches based on quantitative models, the approach here is based on a probabilistic logic model. Assertions about signals are represented by logical propositions, and the model consists mainly of conditional probabilities of the form P( effects ∣ causes ). These probabilities are the results of a statistical evaluation, which is often available in energy plants. In addition to the probabilistic logic description, the causality graph is used to describe the structure of the process, and to supply criteria for the decomposition of the diagnostic task into subtasks. In this way, the complexity of the diagnostic task is reduced, and the efficiency of its solution increased.