An Intelligent Diagnostic/Prognostic Framework for Automotive Electrical Systems

Abstract

Automotive systems are becoming increasingly dependent on electrical components, computer control, and sensors. It has become extremely critical to detect faults in the electrical system and predict the remaining useful life of failing components. This paper introduces an integrated methodology for monitoring, modeling, data processing, fault diagnosis, and failure prognosis of critical electrical components such as the battery. The enabling technologies include signal processing, sensor selection and placement, selection and extraction of optimum condition indicators, and accurate fault diagnosis and failure prognosis algorithms that are based on both the physics of failure models and Bayesian estimation methods. The proposed architecture is implementable on-board an Electronic Control Unit (ECU) requiring minimum computational resources. Potential benefits include reduction in maintenance costs, improved asset reliability and availability and longer life of critical components.