Design And Implementation Of Railway Signal Equipment Failure Early Warning System Based On Machine Learning

Railway signal equipment mainly includes switch machines,track circuits,and signal machines.At present,the fault early warning method of railway signal equipment is mostly the threshold method.The working environment of railway signal equipment is complex,and the threshold method is easy to be affected by the surrounding environment,which leads to the false alarm.Therefore,the development of a more robust method is of great significance to ensure the safe and efficient operation of railway transportation.In this paper,the deep learning model is used to realize the fault early warning of railway signal equipment,the neural network is used to learn the fault-related features of railway signal equipment operation electrical data,and the attention mechanism is used to dynamically allocate weights for the features so that the model can accurately predict the operation status of railway signal equipment and realize the fault early warning.The specific work is as follows:(1)According to the working characteristics of the switch machine that is non-continuous and will complete the action before the train comes,using the working voltage and current of the switch machine as characteristic data.The neural network is used to automatically extract the features related to the fault in the data,complete the fault prediction,and complete the fault warning of the switch machine based on the prediction result.The experimental results show that the turnout fault early warning method in this paper is effective.(2)According to the continuous working characteristics of the track circuit and the railway signal,by analyzing the electrical characteristic data during its operation,a fault early warning method of the track circuit and the signal machine based on the dynamic threshold prediction method is proposed.Use the Encoder-Decoder structure based on Long and Short-Term Memory Network to predict the subsequent operating status of the equipment.The upper and lower thresholds are set according to the predicted value and the 3σ principle,and the fault warning of the track circuit and the signal machine is completed.The experimental results show that the track circuit and signal machine fault early warning method in this paper is effective.(3)Use the PyQt framework to build a fault early warning system for railway signal equipment.By using the packaged functional modules,users can conveniently use data preprocessing,model training,and model evaluation functions to perform fault warning tasks for railway signal equipment.