Railway foreign body vibration signal detection based on wavelet analysis

Abstract

Based on the wavelet packet analysis method with time-frequency analysis characteristics, the measurement signal of the vibration system is processed for noise reduction, the soft-hard threshold compromise wavelet denoising method used has the advantages of soft threshold and hard threshold denoising, and through the introduction of compromise factors, signal processing can be performed more flexibly in signal analysis. For the denoised signal, the fundamental wavelet time-energy spectrum analysis, the main components of the signal can be clearly displayed, and according to the distribution of its energy in each frequency band, the signal characteristics can be displayed intuitively. Experimental results show: It can be determined that there is a foreign body intrusion incident at a position 520 m away from the monitoring point, rather than a normal train travel incident. In fact, we are walking back and forth at a distance of about 520 m from the monitoring point, simulating the intrusion of illegal foreign objects such as pedestrians and livestock beside the railroad tracks prove that analysis and judgment can be known, the wavelet analysis proposed by the author can realize the monitoring and judgment of some illegal foreign body intrusion incidents such as pedestrians and livestock.