An Optimized Algorithm Based on Information Entropy for System-Level Fault Isolation Strategy

Entropy reduction of dynamic fault isolation system is not only based on the uncertainty induction of candidate components, but on the uncertainty of test information, based on which, dynamic entropy approach in fault isolation is proposed. In each step of optimizing fault isolation, such test should be selected that would make the maximum entropy reduction of dynamic fault isolation system, which is named as the test´s dynamic entropy. Experiment result has indicated that selecting test based on this approach could construct a optimized fault tree and have achieved an effective, fast fault isolation.