Bayesian Online Joint Identification on Axle Positions, Axle Loads, and Bridge Structural Parameters Using Evolving Virtual Axle Configurations Without Axle Detector

Vehicle axle load information is of high importance for the assessment of bridge structures. It includes axle positions (vehicle position plus axle configuration) and axle loads. However, axle positions, axle loads of vehicles, and structural parameters of bridges are all required to be determined, resulting in a large number of unknowns and inducing big challenges. In this study, a novel method for online joint identification of axle positions, axle loads, and bridge structural parameters is proposed with evolving virtual axle configurations. With the concept of virtual axles, the joint estimation problem is transformed into two stages: (1) axle position, axle load estimation, and structural identification given a virtual axle configuration and (2) optimal selection of the virtual axle configuration. This avoids simultaneous estimation of a large number of unknowns. In this process, a recursive strategy is developed such that the virtual axle configuration can be evolved according to the structural response data, allowing for online joint estimation. Illustrated examples for bridges under the passage of different types of trucks with different traffic conditions demonstrate the feasibility of the proposed method for axle detection, axle load estimation, and bridge structural identification.