Performance Degradation Based Reliability Prediction Method For CTCS On-board Equipment

High-speed railway is a key carrier for people to travel,and consequences of failures and events at any level should be considered in the design,development,operation and other life cycle.Reliability of the train control system is necessary to ensure efficient operation of the train,its evaluation and prediction is an important means in reducing failure.Reliability assessment and prediction of the train control system is an important approach to detect,confirm these potential risks and deal with the accidents timely.In the research field of the train control system reliability,most of the researches focus on the reliability evaluation,while reliability prediction is still in the initial stage,and a suitable method system has not been formed yet.The contents and innovations of the thesis are as follows:(1)This paper applied performance degradation theory to CTCS on-board equipment’s reliability.It analyzes fault information firstly,studies the evolution of fault information,and then models performance degradation process with several algorithms,and predicts CTCS equipment’s reliability of future time based on the regular that has been obtained by time series analysis and regression analysis.(2)The most representative and descriptive fault feature information of equipment is extracted,the real fault data of different types is classified and summarized,a set of reliability prediction methods which is suitable for the train control on-board equipment is proposed.Software of reliability prediction is designed and compiled based on performance degradation analysis methods,which characteristic information is loaded firstly,and reliability prediction algorithm is called to predict the equipment reliability in the future.(3)Failure rates of each type fault for different types of trains are predicted by time series analysis,and then an improved time series analysis method is proposed,and predicted effect of the two methods is compared.Results show that the improved method is more suitable for train control equipment.Failure rates of the entire on-board equipment are predicted by regression analysis method.(4)Reliability evaluation index of train control equipment is set up,and the severity of the predicted failure is calculated by FMECA,combining with the failure rate to evaluate on-board equipment reliability in the future time,thus completing goals of forecasting the train control equipment reliability.Degradation process models are established,the reliability of different models is predicted,and the effect is tested and verified with real fault data.The results show that the fitting errors between the real value and the predicted value of failure rate with the proposed reliability prediction method system are small,and the methods can provide a better solution of predicting and describing equipment performance degradation process and reliability changes of future time.It has certain value in the train control equipment reliability prediction.