Vehicle Axle Load Identification Using Extracted Bridge Influence Line via Updated Static Component Technique

Abstract

Bridge weigh-in-motion or moving force identification systems have been developed to screen the heavy truck or monitor its gross weight and axle loads. Bridge surface roughness has been considered a very sensitive parameter to the identification error. This paper presents the algorithm to accurately identify static axle weights by modifying the identification process to include the measured bridge influence line containing the actual road profile. The existing iterative calculation called the updated static component (USC) technique is also utilized to improve the dynamic axle load accuracy. The extracted influence line is obtained from a low-speed test using a known axle weight truck. Therefore, the characteristics of the road roughness and the measurement noise are included in the bridge responses. The effectiveness of the proposed technique is investigated through the numerical simulation and the experiment using scaled models. The results reveal that the identified axle loads become more accurate than those identified using the USC and the conventional regularized least squares methods. The proposed technique effectively decreases the identification errors of moving axle loads on the rough surface with a high measurement noise level. Moreover, the regularization parameter can be easily assigned with a broader range to achieve accurate identification results.