Investigation of discrete wavelet transform for signal de-noising in weight-in-motion system

Discrete wavelet transform for signal de-noising to enhance the accuracy of fiber-optic Weight-In-Motion(WIM) system is presented in this paper. The fiber-optic WIM system working principle is demonstrated and the characteristic of the axle weight signal is analyzed. The algorithm of the discrete wavelet transformation and its implementation is discussed in detail. Through multi-scale decomposition and reconstruction, the classification and feature extraction from vibration signals is realized. Field experiments are conducted to evaluate the performance of the proposed method, simultaneously. Experimental results show that discrete wavelet transform for signal de-noising in weight-in-motion(WIM) system is effective.