Study On Failure Law And Maintenance Strategy Of Railway Wagon Couplers Considering Truncated And Censored Data

Under the new situation that China is vigorously promoting the "highway to railway",railway freight transportation has ushered in the historic opportunity of heavy load and high-speed development.With the increase in the tonnage of railway wagons and the increase in the number of vehicle turnovers,the operating conditions of the coupler device,which is the main load-bearing component have seriously deteriorated.Cracks and fractures on the coupler body and coupler knuckle have increased significantly,which poses a huge hidden danger to the operational safety and the stability of transportation for railway wagons.Simultaneously,due to the lack of exploration of the failure law of couplers under new operating conditions,it is impossible to optimize maintenance strategy of couplers.Therefore,this thesis collects the actual maintenance data to develop the research on the failure law and maintenance strategy of couplers.Firstly,the reliability model of the coupler has been built.Based on the left-truncated and fully-censored data,for the situation that the existing parameter estimation method is difficult to apply the likelihood function solution in the thesis,a hierarchical grid(HG)algorithm is proposed to estimate the model parameters inspired by the gradient descent method and the Nelder-Mead simplex algorithm.An improved expectation-maximization algorithm is also adapted to validate effectiveness of the HG algorithm.The results show that the HG algorithm can effectively reduce the number of iterations and improve the robustness of the algorithm to the initial value.Secondly,the confidence interval of the model parameter has been obtained.Combined with the data scene,nonparametric random-weight-Bootstrap sampling is used to generate Bootstrap samples.The confidence intervals corresponding to different quantile methods of the nonparametric Bootstrap are compared in terms of coverage probabilities,it is found that the coverage probability solved by the BCa quantile method is closest to the nominal confidence level.Based on the solution results of the model parameters,the Weibull distribution is selected as the life distribution of the coupler according to the AIC and the failure mechanism.Thirdly,a preventive maintenance strategy for single-component of the coupler based on reliability has been formulated.A two-level imperfect maintenance according to the current maintenance method and maintenance scope of the coupler is defined.Based on these,a mixed failure rate evolution model is constructed,and the cost-effectiveness method is used to decide the specific maintenance method at each maintenance timing.Compared with the current periodic maintenance strategy,the results show that the proposed maintenance model can reduce the two maintenances of the 16-type coupler body in one replacement cycle,and reduce the maintenance cost of 0.341 yuan per day under the premise of higher operational reliability.Finally,the preventive opportunistic maintenance strategy of the 16-type coupler has been achieved.According to the maintenance method and maintenance frequency obtained by the single-component maintenance strategy,the economic relevance of the coupler body and coupler knuckle is considered to establish an opportunistic maintenance time window.Compared with implementing a single-component maintenance strategy separately,the proposed maintenance strategy can reduce the 7 outages on the basis of higher operational reliability,with a maintenance saving rate of 14.6 percent,which reduces the overall maintenance cost of the coupler by 2.42 percent.