Bayesian probabilistic assessment of occupant comfort of high-rise structures based on structural health monitoring data

Comfort performance of high-rise structures during strong winds is significant to habitants. Despite the significance, procedures for evaluating occupant comfort in serviceability limit states have not been as well developed as those for strength-based design of high-rise structures. One of the difficulties arises from uncertainties associated with the parameters in occupant comfort assessment, which pertain to the acceleration response magnitude and its relationship to human reaction to the motion. The comfort assessment is in general conducted by examining whether the wind-induced acceleration response satisfies some occupant comfort criteria. Such a deterministic approach, however, fails to account for uncertainty inherent in the wind-induced acceleration response as it is affected by the wind field of stochastic nature and uncertainty about the aerodynamic loads and the structure's dynamic behavior. In view of this, a Bayesian probabilistic approach is proposed in this study to evaluate the occupant comfort of high-rise structures. First, a Bayesian regression model is formulated for characterizing wind-induced acceleration responses of a structure by use of structural health monitoring (SHM) data acquired during strong winds, thereby enabling to account for the uncertainty contained in the monitored acceleration responses and quantify the uncertainty in modeling and prediction. Based on the predicted acceleration distribution and reliability theory, a safety index is then elicited to perform probabilistic assessment of occupant comfort in wind-induced motion of the structure. In the case study, field monitoring data acquired from a supertall structure of 600 m high during six tropical cyclones are used to illustrate the proposed approach, including the evaluation of occupant comfort of the structure under extreme wind speeds.