Complexity metrics for auto fault diagnosis based on information entropy

Abstract

Due to plenty of advanced and electronic technology used in the modern auto, the modern auto maintenance workload should consider the influence of the fault diagnosis complexity on maintenance man hour, so it is necessary and valuable to evaluate the complexity of the auto fault diagnosis. In this paper, the method of complexity assessment of auto fault diagnosis is discussed by using the information entropy. Firstly, the theory of complexity measurement based on information entropy is introduced, and then the complexity assessment index and the calculation process are shown. Finally, take the engine overheating fault for example, the hierarchy structure of this fault diagnosis is analyzed, and the complexity value of this fault diagnosis based on information entropy is obtained. The results show that this assessment method has a certain practicality and feasibility, which could be utilized to optimize the auto fault diagnosis strategy and maintenance work arrangement.