A genetic algorithm framework for seismic retrofit of building portfolios to enhance community resilience

Abstract

Earthquakes can impose significant human, societal, and economic burdens, emphasizing the necessity of community preparedness, including mitigation. Seismic retrofitting, a key preparedness measure, seeks to enhance structural resilience against earthquakes. Despite the effectiveness of community-level plans, challenges persist in quantifying their effectiveness and optimizing resource allocation. To address these challenges, this paper introduces a genetic algorithm (GA) framework designed to optimize seismic retrofitting of building portfolios in communities, aiming at achieving community-level resilience goals. The proposed framework, part of a simulation-based approach, probabilistically models building damage and post-earthquake functionality of the community. It calculates seismic retrofit levels (in terms of percentage shift in the median of fragility curves) for each building archetype, considering different shifts in the damage fragility curves and the percentage of retrofitted buildings for each retrofit level. The optimization problem maximizes average post-earthquake functionality, constrained by available community funds. The proposed GA framework is illustrated using the real-world example of Salt Lake City, Utah, in collaboration with the NIST Center for Risk-based Community Resilience Planning, showcasing its application through the Interdependent Networked Community Resilience Modeling Environment (IN-CORE). The results of the considered illustrative example reveal substantial improvements in community functionality under different budget scenarios and retrofit levels. The GA framework represents a significant advancement in supporting informed decisions for community resilience. Community leaders can use its outputs to make informed decisions, considering specific factors and trade-offs between benefits and costs.