PHM-based Fault Identification for Electronics-rich Systems under Uncertainty

摘要

Electronics are essential to most current systems and equipments espe cially for electronics-rich systems. Prognostics and health management (PHM) is in troduced to cope with the essentiality. Due to the complex structures, numerous pa rameters as well as various intangible and uncertainty factors, fault diagnosis based on PHM for electronic systems is a highly desirable but also difficult work. There is relatively rare research-concerning applicable and efficient fault identification for electronic systems. This paper proposes a PHM-based diagnostic Bayesian network approach to perform available and efficient fault identification for electronics-rich system. Compare with other methods, it can deal with either single or multiple fault identification with incomplete information and uncertainty. For illustrate purpose, a numerical example is conducted to apply the proposed diagnostics to electronics with uncertain information. The diagnosis results validate that it is accurate in in complete data conditions even with uncertain fault symptoms. It demonstrates the advantages of the approach to solve the PHM-based fault identification problem for electronics-rich system efficiently.