Genetic Algorithm-based Optimization Model for Emergency Shelters Location Allocation in Public Health Emergency Situations

Abstract

Deploying emergency resources after a disaster is crucial for reducing the event’s impact. A multi-objective (MO) resource allocation (RA) method is developed to achieve efficient distribution of relief items and the best choice of transportation routes, taking into account the variable and persistent nature of rescue operations.  The MO Cellular Genetical Approach (MOC-GA) addresses the concept by incorporating auxiliary populations and neighborhood architecture into the cellular automaton.  The comparative experiment conclusively demonstrates that MOC-GA performs favorably compared to Particle Swarm Optimisation (PSO), Ant Colony Optimisation (ACO), and Dragon Fly Optimisation (DFO) in several metrics such as the Pareto Front (PF), hypervolume, mean objective function value, and PF ratios.  The findings demonstrate that MOC-GA effectively addresses the MO dynamical emergency RA approach, offering decision-makers a more comprehensive range of superior and diversified candidate rescue strategies than alternative methods.  This research analyzes the existing rescue methods and proposes a theoretical rescue strategy to aid decision-makers in making scientifically informed decisions.