Abstract
When a train runs on a bridge, the natural frequency of the bridge decreases apparently due to the effect of the mass of the train. However, there have been no researches that discussed this effect based on actual measurement data. In this study, the authors formulate a TV-VAR model in which the VAR coefficient of a vector autoregression model changes with time stochastically, in order to evaluate the decline in the apparent natural frequency of a bridge at the time of train passing, and then develop an estimation method for unknown parameters based on the hierarchical Bayesian method. By applying numerical calculation results and the actual acceleration response of the bridge at the time of train passing, it was found that (1) the proposed method can evaluate the change in natural frequency accurately, (2) apparent natural frequency decreases actually in the bridge at the time of train passing, and (3) the decrease amount is 14% on average, 20% at a maximum for the target bridge of this study.
