ABSTRACT
Many researchers have developed the bridge weigh-in-motion (BWIM) techniques to estimate axle weight and gross vehicle weight (GVW) of the traveling vehicles. BWIM assumes that the bridge response due to traveling vehicles results from the influence line and the vehicle specification, such as the axle spacing and weight. However, traveling vehicles also cause dynamic responses. In this study, the dynamic response is considered in the observation error with autocorrelation in the time domain. Specifically, it is considered in the covariance matrix of the observation error. The accuracy of three methods, Mehtod-1A; ordinary least square method (OLS), Method-1B; OLS with low-pass filtering of displacement response, Method-2; proposed method considering autocorrelation in bridge dynamic response, are compared. Methods 1B and 2 contribute to enhanced estimation accuracy; however, it is worth noting that Method 1B introduces biases in the estimates due to filtering.
