Multi-objective optimization approach of shelter location with maximum equity: an empirical study in Xin Jiekou district of Nanjing, China

Abstract

Based on the inadequacies and neglect of the equity of refuge resources, refuge demands and the evacuation allocation of traditional methodologies, this study put forwards the multi-objectives layout optimization model of shelters which firstly realizes the maximum equity of shelter location. Our approach has the objectives of maximizing equity, minimizing overall egress time and minimizing the quantity of new shelters. The high-precision population is established through mobile signaling data, while the optimization model adopts a circular circulatory allocation rule derived from a gravity model. The shorter the evacuation time, the larger the shelter capacity and thus more refugees are allocated to the shelter. The evacuation time is determined by the application programming interface (API) of the Baidu Map open platform with Python, which exhibits the authentic evacuation paths and real-time traffic conditions. This study designs a three-stage algorithm ‘genetic algorithm-exhaustive method-evaluation’. The first process of algorithm calculates the minimum quantity of new shelters; the second process selects the feasible layout schemes and determines Pareto optimum solutions; and the third stage evaluates the Pareto optimum solution based on the shelter construction cost and the accessibility from shelters to emergency supply storage points to determine the best location scheme. This study regards Xin Jiekou district in Nanjing as a case area to demonstrate reliability and availability of the proposed methodology.