Experimental Calibration and Validation of Fault Diagnosis and Prognosis Algorithms for Automotive Electric Power Generation and Storage System

Abstract

This paper is a continuation of the work on the model-based fault diagnosis for the automotive electric power generation system (EPGS) presented in [5]. Based on the previous work on the subject, a new and optimized fault diagnosis algorithm for the EPGS system is developed. In this paper, the thresholds for the diagnosis algorithm are selected and calibrated based on experimental data. The test bench used for the calibration and validation process is discussed. Finally the effectiveness of the fault diagnosis algorithm and threshold selection is validated using experimental data collected from the test bench.