Multi-component system maintenance optimization of a rail transit train based on opportunistic correlations

Abstract

Abstract In order to deal with the problems of insufficient or excessive maintenance in the current maintenance activities of China transit trains, this paper develops a novel multi-component system maintenance optimization approach based on an opportunistic correlation model. Based on the minimal reliability and failure rate change rule of each train component, the novel proposed maintenance optimization benefits from an improved opportunistic maintenance model with system structure correlation, fault correlation and reliability correlation under imperfect maintenance. Then, different maintenance modes can be determined by a proposed maintenance factor under the different conditions of components. Specifically, the reliability threshold of each component is also considered to optimize the maintenance cost by the system reliability and operational availability of the train. Furthermore, as the mentioned problem belongs to the NP-Hard optimization problems, a modified particle swarm optimization (PSO) with the improvement of inertia weight is proposed to cope with the optimization problem. Based on a specific case under the practical recorded failure data, the analysis shows that the proposed model and approach can effectively cut the maintenance cost.