Fair UAV emergency supply deployment based on an improved genetic algorithm

Unmanned aerial vehicles (UAVs or drones) are widely used in postdisaster material deployment as professional equipment for emergency relief systems. When emergency supplies are deployed to disaster-affected sites, not only the cost and efficiency of rescue should be considered, but also the fairness of material deployment should be given attention. Based on the consideration of fairness, this study investigates the problem of drone emergency supply deployment with dynamic energy consumption constraints. Relative deprivation cost is also used to measure the difference in the psychological trauma of victims at each disaster-affected site to ensure the fairness of supply deployment. A mix integer programming (MIP) model is established, and an improved genetic algorithm, HRGA, is designed to minimize the cost of drone distribution. A heuristic rule is proposed to find an appropriate initial solution quickly by combining the characteristics of the model-related constraints. Compared with the accurate solution of small-scale studies obtained by CPLEX, HRGA reduces the solution time by 68.76% on average while ensuring the accuracy. Compared with the traditional GA, HRGA has certain advantages in solving quality compared with traditional GA with a maximum improvement of 11.55%, and the experimental results verify the feasibility and effectiveness of HRGA. Moreover, the research results can provide useful references for building a fair and efficient drone emergency rescue system.