Skip to main content
eScholarship
Open Access Publications from the University of California

Post-Earthquake Performance Assessment and Decision-Making for Tall Buildings: Integrating Statistical Modeling, Machine Learning, Stochastic Simulation and Optimization

  • Author(s): Zhang, Yu
  • Advisor(s): Burton, Henry V
  • et al.
Abstract

With the embrace of the seismic resilience concept as a measure of the ability of constructed facilities and communities to contain the effects of an earthquake and achieve a timely recovery, the critical role of tall buildings in supporting community functionality, has been brought to the forefront. This study presents a series of frameworks for performing post-earthquake assessment and optimal decision-making for tall buildings. A machine learning framework to assess structural safety is first proposed and applied to a low-rise frame building. A similar methodology is then adapted for tall buildings, while incorporating robust techniques to deal with the high-dimension feature space that arises because of the large number of structural components. Seismic risk assessment is then carried out for the tall building by comparing the time-dependent probability of exceeding various response demand limits over a pre-defined period considering both mainshock-only and mainshock-aftershock hazard. The risk-based consistency of the limit state acceptance criteria for engineering demand parameters is also examined. Finally, a methodology to support optimal decision-making following the mainshock is developed with the goal of minimizing expected financial losses and, at the same time, ensuring life safety. The proposed prediction model, risk assessment methodology and optimal decision-making strategy can provide critical insights into the seismic performance of mainshock-damaged tall buildings and inform the post-earthquake actions of occupants, structural engineers, insurance companies and policymakers.

Main Content
Current View