Research on signal processing for bridge health monitoring

Under the complex field conditions of the bridge health monitoring system, the interaction of various factors would cause burring or point abrupt or data loss for data acquisition. Therefore, the acquisition signal must be processed to ensure the accuracy of the analysis. Firstly, this paper analyzes and summarizes the problems for engineering data of acquisition system, classifies and summarizes some existing solutions to those problems. Secondly, a signal processing method for bridge health monitoring system is proposed. In this method, in order to acquire the characteristics of the real signal, the least square method is used to smooth the signal, the wavelet transformation is employed to identify the signal singular points from the monitoring data. The reliability of the singular points data is analyzed and diagnosed. The signal characteristic of the real data is interpolated by real data around the singular point if it is diagnosed to be burring point. Finally, this paper uses some relevant results to prove the effect of these methods.