Enhancing emergency medical service location model for spatial accessibility and equity under random demand and travel time

This paper proposes a novel emergency medical service (EMS) location model that considers spatial accessibility (SA), equity, cost, random demand, and random travel time of EMS system. We first construct a utility function that incorporates overall SA and equity. Subsequently, under the constraint of a cost budget, we propose a distributionally robust optimization model with the objective of maximizing the utility. Building upon the Wasserstein ambiguity set, we reformulate the original model as a mixed integer p-order cone programming. To handle the computational challenges posed by norms and data size, we propose a utility cut and lift-polyhedron approximation cut generation algorithm. In the numerical experiments section, algorithm comparison, sensitivity analysis, and model comparison demonstrate the significant advantages of the proposed algorithm and model over different benchmarks and corresponding management insights are provided.