The intelligent fault diagnosis for composite systems based on machine learning

Nowadays, electronic devices are getting more complex, which make it also more difficult to use a single reasoning technique to meet the demands of the fault diagnosis. Integrating two or more reasoning techniques becomes a trend in developing intelligent diagnosis. In this paper we discuss the intelligent diagnosis problems and propose a diagnosis architecture for composite systems, which combines rule-based diagnosis and model-based diagnosis. These two diagnosis programs not only work efficiently with machine learning in different stages of the fault diagnosis process, but also efficiently improve the process by making the best use of their individual advantages.