System diagnosis using a bayesian method

Today’s engineering systems have become increasingly more complex. This makes fault diagnosis a more challenging task in industry and therefore a significant amount of research has been undertaken on developing fault diagnostic methodologies. So far there already exist a variety of diagnostic methods, from qualitative to quantitative. However, no methods have considered multi-component degradation when diagnosing faults at the system level. For example, from the point a new aircraft takes off for the first time all of its components start to degrade, and yet in previous studies it is presumed that apart from the faulty component, other components in the system are operating in a healthy state. This thesis makes a contribution through the development of an experimental fuel rig to produce high quality data of multi-component degradation and a probabilistic framework based on the Bayesian method to diagnose faults in a system with considering multi-component degradation. The proposed method is implemented on the fuel rig data which illustrates the applicability of the proposed method and the diagnostic results are compared with the neural network method in order to show the capabilities and imperfections of the proposed method.