Study On Model Test And Extraction Method Of Surface Pulsating Pressure On High Speed Train

The pulsating pressure on the surface of high-speed trains is the source of aerodynamic noise at the low Mach number.Aerodynamic noise will not only produce serious pollution to the environment around the railway,increase the energy consumption of the railway transportation,affect the lives of the residents along the route,but also produce the vibration on the vehicle body,affecting passenger's comfort and the safety of train operation.And the real train test in the route is affected by multiple factors such as sensors and environment,which not only leads to difficulty in testing,but also the signal to noise ratio of the measured signal is low,and it is difficult to extract the pulsating pressure effectively.Therefore,this paper adopts a combination of model test and modern signal analysis processing technology to study the extraction method of surface pulsating pressure,and provides theoretical guidance for reducing the harm of aerodynamic noise.In this paper,the wind tunnel test rig is used to collect pressure,noise,vibration,and temperature data in combination with pressure,noise,vibration,and temperature sensors and data collectors.According to the high-speed train model test,the use of EEMD-PCA-FastICA algorithm to achieve the blind separation of fan noise and aerodynamic noise,so that the single-channel underdetermined problem is transformed into a positive definite problem,thus obtaining the aerodynamic noise and fan noise waveform characteristics.And the blind separation results show that the estimated number of source signals is consistent with the experimental conditions.The estimated main frequency of the fan noise source signal and the original source signal are both 120 Hz,and the correlation coefficient is 0.763,which is a strong correlation,indicating that the algorithm has achieved an ideal effect.To against various influencing factors in the experiment,signal analysis and processing methods were used to control and reduce different interference factors.This paper proposes the use of Improved Wavelet Threshold Denoising to reduce the effect of electromagnetic and other random disturbances.After denoising,the static surface pressure fluctuates around the atmospheric pressure,which can reduce the influence of electromagnetic and other random disturbances.Meanwhile,design and implement of piezoresistive pressure sensor temperature calibration experiment was conducted to fit the model of the error and temperature between the output value and the actual value of the 8515C-15 piezoresistive pressure sensor.The model shows that when the temperature is kept constant,the effect on the pressure is a certain value,so as to compensate the temperature of the sensor.And the box vibration experiment was conducted to research the effect of vibration acceleration on the output of the pressure sensor.The magnitude of pressure amplitude caused by vibration acceleration is quantitatively described by linear regression.This formula shows the vibration acceleration was consistent with the pressure's change trend,and it can be applied to reduce the vibration of the sensor.Based on the Ensemble Empirical Mode Decomposition,using the correlation coefficient method to reduce the influence of fan noise on the pressure,and finally the extracted surface pressure of the high-speed train model test is approximately ±50Pa.The aerodynamic noise spectrum of the Blind Source Separation is used as a basis and compared with the extracted pulsating pressure spectrum,getting the information that the frequency of these two is the same,and the large amplitude is mainly concentrated in the range of 0-30 Hz,which is correspond with the characteristics of the fluctuating pressure that mainly distribute in low frequency band,and it has verified the effectiveness and correctness of the pulsating pressure extraction method in this paper.The above extraction method is used to extract the surface pulsating pressure of the actual route of the high-speed train,and a pulsating pressure level model is established.In this formula,its amplitude decreases with the frequency of negative index trend,and it can be used to predict the aerodynamic noise of high-speed trains,to optimize the damping and noise reduction structure of high-speed trains,which could provide data support for vibration damping and noise reduction of high-speed trains.