Research On Fault Diagnosis Of Rail Transit Vehicle Speed Measurement And Positioning Equipment Based On Data Drive

Rail transit speed measurement and positioning system equipment is responsible for the calculation and monitoring of train speed and position to ensure the safety and efficiency of trains in actual operation.At the same time,the high frequency of equipment operation caused failures,which greatly affected the operating efficiency of the train.At present,the study on the fault diagnosis of rail transit speed measurement and positioning system equipment is mainly to analyze the text data information of train operation,while the collection and analysis are less.In addition,most of the fault diagnosis of rail transit speed measurement and positioning system equipment relies on the experience of experts and technicians,which leads to a substantial increase in the workload of train equipment maintenance personnel and reduces maintenance efficiency.This thesis developed a software named data receiving system,targeted collection of equipment operation information of rail transit speed measurement and positioning system,and used the association rule Eclat algorithm to analyze the fault log of rail transit speed measurement and positioning equipment.The CNN-LSTM neural network model is used to classify and extract the spatial and temporal characteristics of the fault data information of rail transit speed measurement and positioning equipment,so as to realize the fault diagnosis of rail transit speed measurement and positioning system equipment.The main study contents of the thesis are as follows:(1)Development human-computer interaction log receiving system software,receive real-time information about the operating status of rail transit speed measurement and positioning system equipment,and store it according to the needs of fault diagnosis in the thesis,and provide a data basis for the fault diagnosis model of this thesis.For simple obvious equipment failures,it can be directly judged through the display interface of the log receiving system software,which saves the time for fault diagnosis.(2)Comprehensive introduction and analysis of rail transit testing and positioning system equipment,determining the type of faults studied in this thesis,and analyzing the fault log of rail transit speed measurement and positioning equipment through the association rule Eclat algorithm to extract data that has strong correlation with rail transit speed measurement and positioning equipment Item,which provides input for the fault diagnosis model.(3)Based on the advantages of CNN neural network model and LSTM neural network model in processing fault data information in space and time respectively,these two models are selected as diagnostic models and introduced.Besides CNN-LSTM neural network fault diagnostic model are constructed according to their network structure.(4)Establish a CNN-LSTM neural network model,optimize the CNN-LSTM neural network model by using Adam algorithm training,select the optimal batch data volume,and use correlation coefficients and accuracy to verify the feasibility of the CNN-LSTM neural network model in the circumstances of optimal batch data volume.