Optimizing Humanitarian Relief: A Hybrid MIP-NSGA-III Framework

Humanitarian logistics forms the backbone of disaster response efforts, as its efficiency is a decisive factor for the survival and well-being of affected communities. However, this field faces multiple challenges, including information asymmetry, resource constraints, and coordination difficulties, particularly in volunteer scheduling and task assignment. Traditional decision-making methods often struggle to cope with the complex and dynamic relief environment. This research introduces a multi-objective optimization scheme for volunteer task allocation in humanitarian relief. Leveraging a mixed-integer programming framework and the Non-dominated Sorting Genetic Algorithm III (NSGA-III), the framework concurrently optimizes three conflicting objectives: minimizing logistics expenditure, maximizing the fulfillment of beneficiary needs, and enhancing the efficiency of personnel transfer within evacuation networks. By constructing a two-tier network model comprising suppliers, relief camps, affected areas, and evacuation camps, and incorporating uncertainty factors such as camp disruptions and route risks, this research strives to more accurately reflect the complexity of humanitarian relief operations. The performance of the proposed NSGA-III model is assessed against the Multi-objective Evolutionary Algorithm based on Decomposition (MOEA/D), a widely used multi-objective evolutionary optimization scheme. Numerical experiments confirm that NSGA-III effectively resolves this complex multi-objective problem, yielding a well-distributed and diverse set of Pareto-optimal solutions. This outcome equips decision-makers with a comprehensive variety of options for navigating trade-offs according to their priorities. In a comparative analysis, NSGA-III shows improved performance over MOEA/D with respect to diversity and convergence, especially within high-dimensional objective spaces. Its reference point-based mechanism proves more adept at preserving a spread of solutions. Consequently, this research offers a robust decision-support tool for optimizing volunteer management and resource allocation in humanitarian crises. By improving the efficacy and operational performance of relief efforts, this work contributes valuable insights to both the theoretical and practical advancement of humanitarian logistics.