Reliability Analysis Of Electric Multiple Unit(EMU) Subsystem Based On Maintenance And Research On Its Maintenance Period Decision Model

Reliability analysis of EMU subsystem and optimum research on its maintenance strategy have become a key issue to secure safe and efficient operations of EMU. Based on EMU failure and maintenance data from investigation, and reliability theory and models from summarization and analysis, the paper classifies the EMU failure levels, analyzes the criticality of failure model of EMU sub-system and its life-span distribution, researches on the influence of preventive maintenance to reliability to EMU sub-system and the optimization of periodic decision-making model of preventive maintenance, and finally draws the conclusions as following:(1) After conducting statistical analysis of existing fault data and making advantage of analytic hierarchy process, the fault of EMU is divided into four grades based on three dimensions including influence level, influence time and the probability of occurrence. Since the evaluation index is targeted at the faults themselves, the conclusions drawn have a certain significance of analyzing faults and optimizing maintenance strategies. Carrying out statistical analysis of high-voltage traction-driven system for a certain type of EMU, and putting forward a calculation method of fault mode severity combined with fault grades, through which severity of various common fault modes is calculated. Moreover, corresponding measures are proposed targeted at fault mode with high severity to improve the reliability of components prone to such fault modes.(2) Reliability function of EMU under the condition of maintenance and expression of mean time to failure are derived; Adopting statistics and reliability theory obtains the fitting result of the basic life distribution function of high-voltage traction-driven system of a certain EMU type, and verifies that its distribution conforms to Weibull distribution. The qualitative and quantitative comparative analysis is conducted after the introduction of a new reliability function model under preventive maintenance, which shows that preventive maintenance can significantly increase the service life of this system and can slow down its reliability’s declining speed. Meanwhile, it also verifies that the existing preventive maintenance period of this system is relatively more reasonable.(3) A model is put forward to employ Monte Carlo Method applicable to the objective of maximum availability developing preventive maintenance period. Adopting a multi-objective optimization method optimizes the proposed model and set up a comprehensive evaluation index comprehensive considering availability, economy and reliability of the EMU in a certain repair period. This model can change the priorities of availability, economy and reliability in calculation process by adjusting the weight coefficient, and boasts a wide range of applications. The optimum preventive maintenance period of high-voltage traction-driven system of a certain EMU type as an example analyzed of the both model, and gets preventive maintenance cycle of this sub-system under each certain objectives.