Investigation of fault detection and analysis methods for central maintenance systems

Abstract

Modern airplanes have a large number of interconnected avionic systems that need an accurate and confident real-time fault detection/isolation (FDI) monitor in central maintenance systems (CMS) for possible malfunctions or abnormalities. Model-based diagnosis, which compares predicted behaviour with actual behaviour, is used for efficient analysis of maintenance data for FDI. This paper investigates the different types of model-based fault detection/ isolation approaches according to type of aircraft. The importance of real-time (in-flight) diagnosis is discussed.