ABSTRACT
The first step in a BWIM system is to obtain the speed of the vehicle to be weighted. There exist several possibilities to estimate the vehicle’s velocity, being the most popular the method that calculates the correlation between signals from two separate sensors. In practice, this method is sufficiently accurate for most BWIM installations. However, it does not perform well in certain situations, most notably in the case of an ideal simply supported configuration or for systems with highly asymmetric influence lines. This paper introduces a new theoretically consistent method, termed convoluted reciprocity, that allows for accurate vehicle speed estimates for any structural configuration and shape of influence line. This document presents in detail the theoretical background of the novel approach complemented by numerical simulations. The performance in speed estimation of the new method is compared directly to the estimates obtained via the correlation method. In addition, these methods are compared empirically with the measured responses of a scaled bridge under controlled laboratory conditions. The results show that the convoluted reciprocity idea does provide very accurate results, while it relies on the same sensor configuration of standard BWIM systems.
