A Robust Neural Network Algorithm For Automotive Air Conditioning Fault Detection

Abstract

Accurate and incipient fault detection of air conditioning systems is highly demanded in a car to prevent energy waste and high maintenance cost. However, most fault detection techniques require experiences of drivers which are usually unavailable. In this study, a novel hybrid method is proposed to detect faults for AC systems in car. Two typical faults in AC system are adopted to investigate. An AC fault detection and diagnosis framework is introduced by combining the RBFNN model and the EWMA. The results show that the proposing algorithm detects typical air conditioning faults in a car with high accuracy.