Reliability Analysis Of EMU Pantograph System Based On Multi-Source Data Fusion

With the development of high-speed EMU technology,the reliability level of pantograph system,as an important part of EMU traction system,has made a great leap,but due to the constraints of long period and high cost of trainset reliability test,it is difficult to obtain enough failure data within a limited time,and the limited failure data limit the accuracy of component reliability analysis results.On the other hand,a certain amount of reliability data has been accumulated during the operation and maintenance of EMU,but it has not been fully utilized,how to make full use of multi-source data to analyze the reliability of EMU pantograph system is of great significance for improving the reliability level and ensuring the safe operation of EMU.In this thesis a pantograph system reliability analysis method integrating multiple sources of data is proposed based on Bayesian theory,which can improve the accuracy of pantograph system reliability analysis by making full use of multiple sources of reliability data such as pantograph system operation mileage data,expert experience information generated during overhaul,and performance degradation data to study its fault evolution law.The main research contents of this thesis are as follows:(1)The reliability analysis method of the pantograph system by fusing the running mileage data and expert information is proposed.Using for reference the methods of dealing with expert experience information in aerospace and defense industry,the expert experience information is quantified by point estimation and interval estimation of reliability index,it is transformed into a priori information of the pantograph system Bayesian cognitive science,and the likelihood function is constructed with the pantograph system operating mileage data as the life data.Finally,the reliability analysis of the pantograph system is carried out to provide theoretical support for the reliability analysis of the pantograph system under the conditions of small amount of fault data and non-negligible expert experience in the early operation period(2)The reliability analysis method of the pantograph system considering the degeneration of individual difference is constructed.In order to solve the problem of individual difference under the influence of uncertain factors such as machining design,operating circuit and maintenance degree,a reliability modeling method of pantograph system carbon slide plate with individual difference degradation is proposed by using the performance degradation data.A stochastic process model was established by introducing random effect parameters to characterize the performance degradation processes of different individuals,and a performance degradation data likelihood function was constructed based on the properties of the inverse Gauss process,the uniform distribution is chosen as the prior distribution without information,and the posterior parameter is estimated by Bayesian theory,taking the performance degradation data of several carbon slide plates of the same type in the pantograph system of Lanxin high-speed Railway as an example,complete the pantograph system reliability analysis considering individual differences.(3)The reliability analysis method of pantograph system based on multi-source data is constructed.Aiming at the problem of single source of data for pantograph reliability analysis,a reliability analysis framework based on the fusion of multiple sources of data is constructed,and objective data including operational mileage data,performance degradation data and subjective information such as field experts’ experience knowledge are integrated into the reliability model construction process to improve the confidence of reliability analysis.The proposed method is validated by taking the carbon slider of pantograph system as an example,with provides a new idea for the reliability analysis of pantograph system of EMU.