On Vehicle Fault Diagnosis: A Low Complexity Onboard Method

Abstract

Implementing real-time and onboard fault diagnosis on electric vehicles can effectively avoid potential dangers. However, the low calculating ability and limited storage capacity of electric vehicles hamper the development of real-time and onboard fault diagnosis. To address the issue, combining neural network and fuzzy logic, we propose a low complexity onboard vehicle fault diagnosis method to monitor the vehicle status and give early warning of accidents. In twelve months, we first utilize three electric vehicles and collect 6. 52GB real data related to vehicle components. Motivated by those data, we conducted an in-depth research on the major vehicle faults, and divided them into four types which are no fault, battery fault, sensor fault, and module fault. Furthermore, we propose a BP neural network based multiple training method to define the correlation between data types and fault types. Then, applying the correlation and data, a fuzzy logic based classification method is proposed to evaluate the vehicle status and give early warning. Finally, a comprehensive simulation is conducted, which indicates that the accuracy is 88%.