Model-based fault diagnosis using arti?cial intelligence techniques often deals with uncertain knowledge and incomplete information. Probability reasoning is a method to deal with uncertain or incomplete information, and Bayesian network is a tool that brings it into the real world application. A new method is proposed to construct a Bayesian network model based on fault tree and bond graph theory, and diagnosis of system. Realization of localizing faulty system components that cause the abnormal behaviors of a system or process, and at the same time to get the faulty component on the system impact of the size, in other words it provides system operators a priority checking and maintenance schedule for system components.. Finally, using simulation to verify the performance of this method.