Fault Diagnosis and Isolation for Brushless Electronic Returnless Fuel Systems

Abstract

The brushless electronic returnless fuel system (ERFS) is a complex system, which is subject to different faults such as motor resistance increase and pressure sensor bias. Any of these faults may lead to no-start and safety-related issues such as loss of engine power. It’s challenging to develop a diagnosis and isolation method to detect and isolate faults in the field, with feature signals which capture different fault signatures. In this study, we first develop a set of feature signals which are used to monitor the behavior, track the degradation, predict the potential failures, and diagnose the issues of the brushless ERFS. The feature signals include the estimated resistance, current error ratio, PWM duty cycle error ratio, zero pump speed ratio, etc. Then we develop a fault diagnosis and isolation algorithm to generate the diagnostic results based on feature signals. The algorithm is tested on a vehicle with fault injection. The results show that different faults can be identified correctly with the developed algorithm. This study enables development of diagnosis and prognosis for the brushless ERFS, which can protect customers from walk-home scenarios and safety related issues due to the brushless ERFS failures.