Research On Fault Location And Signal Reconstruction Of Train Bearing Based On Microphone Array

The high speed development of railway transportation productivity has put forward higher requirements for traffic safety.The bearing fault of the train wheelset is one of the important reasons for the safety problems of the train.It is of great significance to detect the fault of the wheelset bearing of the train in real time for the safety of the train.The existing fault diagnosis methods for the bearing on the track mostly use the infrared axis temperature detection technique of the base station to diagnose the bearing fault,which has several problems like high false alarm rate or obscure early fault detection.In this paper,we take the sound signal generated by train passing as the research object,and focus on the following two problems:fault location and signal distortion in railway track bearing acoustic diagnosis system,propose a spatial smoothing algorithm based on instantaneous window,locate the wheelset bearings with fault,take a new method of reconstructing sound signals with Dopplerlet transform and instantaneous angle time resampling by matching tracking,and design and realize the scheme of microphone array signal acquisition system.The experimental results show that the new method is accurate in locating the sound source of the fault,and it has a certain correction effect on the distortion of the sound signal.The specific research contents are as follows:(1)The study on the mechanism of sound source positioning of fault bearing.First,the fault mode of the rolling bearing and the frequency of the fault feature are analyzed,and the signal bandwidth produced during the operation of the fault bearing is obtained.Then,the corresponding location method of fault location is investigated,and a MUSIC spatial smoothing algorithm based on instantaneous windowing and frame division is proposed to solve the problem of the movement of the fault source and the coherence between the fault sources.Finally,the relationship between the location parameters and location accuracy is analyzed through simulation,and the effectiveness and accuracy of this method for locating the location of mobile fault are verified.The research of this part provides parameter considerations for the design of microphone array signal acquisition system,and also provides important parameters for time resampling in the reconstruction of fault sound signal.(2)The design of a microphone array signal acquisition system.First,the topology of microphone array is designed,and the microphone selection is investigated,the electret microphone with high sensitivity is selected as the array element.Then the STM32 processor of data acquisition system based on hardware module is designed,which contains a signal amplifying circuit,power supply circuit,high-speed AD sampling circuit and design of Ethernet transmission circuit,realize the synchronous sampling of eight channel signal,the sampling frequency is up to 50Ksps.Finally,a fault acoustic signal acquisition experiment is carried out for the designed location system,the rationality of the acquisition system and the effectiveness of the fault location algorithm are verified by algorithm.(3)The research on the reconstruction of sound signal.When the train is running,the sound signals from the bearing will produce Doppler effect,which leads to the distortion of the fault signal and can not be accurately diagnosed.To solve this problem,a Dopplerlet transformation algorithm based on matching tracking is proposed to revise the signal frequency,and to resample the time point according to the real-time angle of the sound source,and then to calculate the motion parameters of the sound source moving process,and verify it by experiments.The results show that after processing the above algorithm,the location of the bearing can be accurately located,thus correcting the Doppler distortion of the sound source,which plays an important role in the fault diagnosis and feature extraction of the bearing's acoustic signal.