Statistical analysis of the random failure model of high-speed railway equipments

Chinese high-speed railway is undergoing a very vast development in recent years. How to efficiently maintain the high-speed railway system becomes an urgent task faced by the railway operator. In this paper, we study the random failure model of high-speed railway equipments in China, based on the statistical analysis of the real data of failure records. Several classical distribution functions for the random failure model are compared, including the Weibull distribution, normal distribution, and lognormal distribution. Based on the observation of the two different trends on the failure records, we propose a mixed distribution model to estimate the random failure behavior. We find that the mixed Weibull distribution is the best one among three different models. This finding is validated by the value of fitness and D_n in the K-S test. This work provides a more accurate estimation method for the random failure model of high-speed railway equipments, which can be used to further study the maintenance optimization and decision making.