Information Entropy Based System's Diagnostic Parameter Choosing Method with Backward and Forward Reasoning

This paper has introduced a diagnostic parameter choosing method for a multiple-input and multiple-output system. This method uses the entropy of the visible diagnostic parameter set as the criterion to classify the heuristic diagnostic parameter set,and on basis of the subset which is decided by the system's current state,forward reasoning is carried out to get a fault subset, regarding which being the object,backward reasoning is used to decide the final fault set. The element's correspendent measurable diagnostic parameters in the final fault set are the desired system's diagnostic parameters. This method can be a part of a diagnostic expert system.