Light Rail Vehicle Moving Load Identification Based on Wheel Load Allocation

Abstract

In the context of the increasing need for structural health monitoring of public railroad bridges, vehicle loads, as a crucial influencing factor, could be assessed through bridge weigh-in-motion (BWIM) to quantitatively manage their impact on bridge service life. However, most existing assessments focus primarily on straight bridges and tracks, with relatively few studies addressing horizontal curved bridges. This limitation stems from the complex bend–twist coupling characteristics of curved bridges, which complicate deformation and vibration analyses. To address the challenge of predicting moving external loads on horizontal curved bridges, an approach combining the Moses algorithm with a wheel load distribution algorithm was proposed to estimate the weight of each wheel in this paper. To obtain a clear signal as each wheel passes through the measurement point, the strain–time curve was extracted by wavelet transform, and the relevant signals were identified. This enabled the measurement of key vehicle parameters, including the number of axles, vehicle speed, and axle spacing. The results demonstrate that the relative errors between the actual and theoretical values for gross vehicle weight and axle weight are 12.96% and 16.86%, and the relative errors of axle spacing and speed are 9.83% and 1.6%. Notably, the smaller curvature of the curved beam contributes to the smaller difference between the inner and outer wheel loads measured by BWIM.