Research On Reliability Of Differential Protection System Of Electric Multiple Units Trains Based On Monte Carlo Method

With the increasing speed of railway traffic and the increasing proportion of high-speed trains,the operational safety of high-speed trains and the operational reliability of various systems are getting more and more attention,and the corresponding maintenance workload is also followed.Multiplying,if there is no efficient,scientific,and accurate maintenance plan,it will inevitably cause a great waste of manpower,material and financial resources.The differential protection system is an important guarantee for real-time monitoring of the traction main circuit of the EMU.Failure or misoperation will affect the normal operation of the entire EMU,and its reliability will directly affect the safety and stability of train operation.Therefore,the reliability research of the differential protection system of the EMU has great engineering significance for the safe and stable operation of high-speed trains and the long-term stability of railway traffic order.This dissertation takes the EMU differential protection system as the research object,analyzes the failure mode and failure mechanism of the main components of the differential protection system,and constructs the fault tree model of the EMU differential protection system.Using the downward method to calculate the minimum number of cut sets of the fault tree model is 85.First,in view of the feature that the high-reliability EMU differential protection system can only obtain small sample fault data,the support vector regression machine is used to expand the small sample data set.Sufficient and large sample data can ensure the accuracy of reliability research.On the basis of the established fault tree model of the differential protection system of the EMU,collect the truncated fault data of each bottom event and the corresponding prior knowledge,research and analyze the fault data set,and determine that the distribution function of each bottom event is two Parameter Weibull distribution,Bayesian estimation method combined with Markov chain and maximum likelihood estimation method are used to estimate the parameters of the probability distribution function of the bottom event occurrence,and the probability density distribution function of each bottom event is obtained as the EMU differential protection Prepare for the reliability study of the system.Introduce the traditional Monte Carlo simulation method,aiming at the established fault tree model,use the inverse function of the probability density distribution function of each bottom event to perform a large number of sampling to simulate the repeatability experiment,and use the sampling result as the input of the constructed top event Boolean function,Perform curve fitting on the output result to obtain the system failure probability distribution function,draw the system failure distribution curve to find out the change law of the EMU differential protection system failure.And calculated the importance of each event to find the weak link of the system.Finally,the mathematical concept multi-level Monte Carlo method is introduced into the engineering field for reliability simulation analysis,all cut sets are hierarchically nested,and the calculation of partial cut sets is abandoned when the corresponding required accuracy is reached,and the calculation results are found to be consistent with the traditional.The Monte Carlo method is similar and can greatly reduce the calculation time.By calculating the importance of each component to the reliability of the system,the weak link of the differential protection system of the EMU is found.It is found that the winding is the most important to the differential protection system,and the core rust,resonance,external short circuit,insulation damage and other events.The higher the importance of this should be paid attention to in the overhaul.The research results prove that the method proposed in this dissertation has great advantages in the reliability research of large and complex systems such as EMUs,and can meet the needs of current rapid reliability calculations,and is used to formulate scientific maintenance plans,operation and maintenance decisions,and products.The optimized design of the company provides a theoretical basis for reference.