Using Two Sensors for Damage Detection of Bridge Based on Transfer Entropy of Vehicle-bridge Information

Abstract

An innovative damage identification method using only two sensors for damage detection of bridge based on transfer entropy of vehicle- bridge information is proposed in this article. In this method, a test vehicle passing through the bridge surface, and structural vibration acceleration responses at two measuring points are measured from two acceleration sensors, which one set on the bridge, and the other one set on the test vehicle. A fixed moving window is defined as a definitive length calculated by the sampling frequency and the fundamental frequency of the measured acceleration responses. The windowed pair time series extracted from the two measured acceleration responses that obtained from the bridge and the moving vehicle respectively are used to calculate the transfer entropy, which is used to define the new damage index. A simply supported beam bridge subjected to a test vehicle is simulated to verify the effectiveness and feasibility of this method. Numerical simulation results show that this method can accurately detect the damages under both single damage scenarios and multiple damage scenarios using acceleration responses obtained from bridge and moving vehicle. It indicates that the method proposed in this article has a promising application prospect in the field of bridge damage detection.