Research On Health Monitoring Method Of High-speed Train Rolling Bearing Based On Acoustic Emission Technologh

As an important part of rail transportation,high-speed trains can run at speeds of up to 350km/h.Due to its fast running speed,high-speed trains are subjected to more complicated load conditions than ordinary trains.Therefore,higher technical requirements are imposed on the working performance of various components of the train.The coupling bearing between the axle and the steering frame member is one of the important components of the transmission,and the running state of the bearing directly affects the operational safety of the train.If the early failure of the bearing can not be detected in time in the process of high-speed train travel,it may occur serious safety accidents.Therefore,real-time dynamic monitoring of the running state of rolling bearings has important application value in the field of rail transit.In this paper,the rolling bearing of high-speed train is taken as the research object,and the theoretical research and experimental analysis of fault diagnosis theory of bearing acoustic emission waveform flow signal are carried out.Firstly,the fault feature extraction of the acoustic emission waveform stream signal is studied,which is based on signal noise reduction preprocessing and signal envelope extraction.According to the characteristics of acoustic emission waveform stream signal,this paper proposes a parameter-optimized LMS adaptive filtering technique to denoise the acoustic emission waveform stream signal,and select the appropriate reference target signal to complete the adaptive noise reduction of the signal.The validity of the algorithm is verified by the analysis.Then the envelope of the waveform stream signal is extracted.The interpolation method is used to extract the envelope of the fault signal,and the traditional cubic spline interpolation is supernormal in the envelope fitting process.The problem is adjusted,and a subdivision algorithm is proposed to improve the cubic spline interpolation.The simulation results show that the algorithm can effectively extract the envelope of the signal.Then the test data acquisition and analysis of high-speed train rolling bearings are carried out.Based on four kinds of natural fault bearing with different damage(normal,weak,moderate and serious)disassembled from high-speed train,acoustic emission technology is used to monitor and collect signals of four bearings at three representative speeds of 50 km/h,150 km/h and 250 km/h.The experimental data are processed to verify that the proposed algorithm can effectively extract the signals.The fault characteristic frequency can accurately determine the location of bearing fault and complete the qualitative judgment of bearing.Finally,the damage severity of rolling bearing is quantitatively analyzed.In this paper,four relatively stable characteristic parameters(spike factor,spike incidence,RMS and absolute energy)of waveform flow in time domain are used to analyze the trend fitting of 15 groups of test sample data at the same speed.The analysis results show that these four characteristic parameters can effectively evaluate the severity of bearing failure.Then,the accuracy of quantitative evaluation of test bearings is further verified by using the characteristic fingerprint parameters based on impact as auxiliary criteria.Through qualitative diagnosis and quantitative assessment of bearing damage status of high-speed train rolling bearings,it is proved that acoustic emission monitoring technology is an effective and reliable method for fault diagnosis and condition assessment of bearings.