Integration of failure prediction Bayesian networks for complex equipment system

With the advantages of the modularization concept, this paper proposes an integration modeling method of failure prediction Bayesian network (FPBN) for failure prediction of complex equipment system. First of all, the definition of failure prediction Bayesian network module (FPBNM) is introduced and described. Then, when the complex equipment system is decomposed into some subsystems and represented with a set of related FPBN models, the corresponding modularization method of FPBN to FPBNM and the integration method of FPBNM models are discussed in details. Moreover, based on the super node mode of integrated FPBN model, this paper proposes a convenient and efficient inference algorithm. Finally, the case study of FPBN integration for an airplane head up display (HUD) system is carried out. The result shows that the proposed integration method of FPBN could build and inference the practical model efficiently for such complex equipment system.