Bayesian Prediction of Structural Response Under Crowd-Induced Walking Load

Research on modeling the walking-induced dynamic load and its effects on structures has been increasing in recent years. Existing standards for estimating the acceleration response excited by crowd takes [Formula: see text] ([Formula: see text] is the number of walking persons in a crowd) times the single pedestrian load as the crowd load, neglecting uncertainties in the related parameters. This paper proposes a method for predicting the crowd-induced structural acceleration response based on the Bayesian theory. Parameters of the crowd include the distribution of stepping frequency, the interval of arriving time and the weight of pedestrians. By combining the existing methods for calculating the structural response to a pedestrian load and the Bayesian theory, the root-mean-square acceleration along with the crowd parameters are obtained. A field investigation is conducted to study the walking characteristics of pedestrians, including the distribution of stepping frequency, walking velocity and step length. The case of a crowd moving across a simply supported beam bridge is taken for illustration. The acceleration response under the crowd load with different stepping frequencies is calculated. A comparison between the results from the Bayesian method and the design standards is conducted. The proposed method is found to yield acceptable results, and is then applied to estimating the acceleration response of a typical Tibetan heritage building, with the result compared with a field measurement.