A reliability analysis method for complex mechanical systems containing probabilistic-interval information

In this paper, a new reliability analysis method for complex mechanical system which contains hybrid uncertainties is proposed. The hybrid uncertainties containing randomness and intervals are considered in the limit-state function of reliability. Middle Point Limit-State (MPLS) model is first proposed for calculating the reliability index of hybrid uncertainties, and this model could make the index interval smaller and accurate. For complex mechanical systems, system hierarchy technique is used in reliability modeling, and composite limit-state function is then obtained to calculate the system reliability. By constructing the optimization model of the composite limit-state function, the reliability index interval can be obtained, and artificial bee colony algorithm is employed to solving these optimization models. The reliability index interval obtained through the method proposed by this paper could be more precise comparing to existing techniques. The reliability indexes containing the interval and MPLS index may be a more useful analysis tool to assess the mechanical systems reliability with hybrid uncertainties. A numerical case is carried out to illustrate this method and an engineering case of the satellite driving system is used as a typical complex mechanical system to demonstrate the analysis method proposed in this paper.