Reliability Measurement for Multistate Manufacturing Systems with Failure Interaction

Reliability is one of the important factors for manufacturing system. Most researches assume that the failure is independent and the components only have two states, which will lead to inaccurate results. In this paper, a reliability model is proposed considering both failure interaction and multi-state property of the manufacturing system. Starting with a two-component system, a function of state probability under the impact of failure interaction is established after the analysis of failure interaction. Then the multi-component system is decomposed into several subsystems and the failure interaction coefficient is estimated in each subsystem with a Copula function and the Grey model method. Finally, the reliability model is realized with the performance generating function which is derived with the UGF technique and failure interaction coefficients. An example of a cylinder engine manufacturing system is studied, and the result is closer to the practical data.