Identification of Subway Tunnel Vibration Signal Based on Fiber Grating Array

A method is proposed for the identification and classification of surface intrusion events based on the fiber grating sensing network in a subway tunnel. The method combines spectral subtraction and the root mean square of power spectral density to effectively extract event signals, followed by wavelet packet method for feature extraction and SVM for vibration event classification. The proposed scheme is experimentally verified to be able to identify four common events, including continuous pulse construction, discrete pulse construction, lorry passing on the surface, and subway train traveling. The average recognition rate reaches 90.70%, which can satisfy the requirements of practical applications.