Development Of Fault Diagnosis And Monitoring System For High-speed Train Bearings

In recent years,with the continuous development of China's high-speed train industry,ensuring the safety and stability of train operations has become one of the research targets that cannot be ignored by relevant researchers in China.The high-speed train bearing is an important component on the running part of the train,and it is also vulnerable to damage.There are many researches on bearing fault diagnosis nowadays,but few systems are really applied to train bearing fault diagnosis.There are many problems in the maintenance of bearings in the enterprise.Therefore,it is necessary to carry out research on fault diagnosis and inspection.This article first introduces the commonly used axlebox bearings for high-speed trains in China through field research.And then briefly summarizes the common fault types,causes of faults,and the effects of faults on high-speed train axlebox bearings.In this paper,the mechanism of bearing vibration is studied through three-dimensional modeling and modal analysis of the actual axlebox bearings.In this paper,fault signal simulation is used to collect bearing vibration signal data to analyze the vibration signal characteristics of rolling bearing under different fault conditions.This paper introduces fault signal extraction methods of several different fault diagnosis methods in detail.Fault diagnosis methods include time domain index method,spectrum analysis method,time-frequency analysis method and resonance demodulation method.Fault signal extraction methods include time domain index method and wavelet packet.Decomposition energy method and EMD decomposition kurtosis method.Finally,three neural networks were established by using different feature extraction methods,including nine diagnostic models including BP neural network,PNN neural network and RBF neural network.This article uses hybrid programming technology to combine LabVIEW and MATLAB software as the software foundation of programming,combines a variety of signal processing and fault diagnosis methods,designing and building software systems,and applying tools such as data acquisition cards and sensors to build hardware systems.Combined hardware system to establish overall fault monitoring and diagnosis system.In the end,this paper conducts a comprehensive experimental comparative analysis of the system through three different data sources experiments,which fully proves the accuracy and feasibility of the instrument system.This instrument system has a friendly interface,easy operation,and a high total accuracy of the method has a certain application value.