| description abstract | When facing multiple, competing, objectives such as those encountered in resilience-based design of buildings (e.g., cost reduction versus functionality preservation), there is no single design that can be declared objectively better than the rest. Multiobjective optimization (MOO) methods can instead be used to identify a group of design alternatives that represent the optimal trade-offs between competing objectives. In doing so, however, MOO relies on performance evaluation for a large number of design alternatives, representing a problem for performance-based building seismic design, which typically involves computationally intensive and time-consuming nonlinear analyses of building models subjected to suites of ground motion records. Furthermore, only minimal guidance exists for the post-MOO selection of a final design from the optimal set, further impeding its adoption in design of buildings. This paper addresses these issues through use of simplified structural models and methods applied within a MOO methodology, coupled with a newly adapted post-Pareto pruning approach to narrow down the optimal design set to a manageable-sized group of final design alternatives. The methodology was applied to optimize the design of a 7-story reinforced concrete moment frame office building in San Francisco. | |
