The efficiency, equity and effectiveness of location strategies in humanitarian logistics: A robust chance-constrained approach

In the preparedness phase of humanitarian logistics, uncertainties from both the supply and demand sides may dramatically increase morbidity and mortality. We consider a distributionally robust facility location model with chance constraints in which the nodes and edges of the network are vulnerable to random failure. Efficiency, effectiveness and equity metrics, which can be explicitly demonstrated as operational costs, service quality and the coverage rate, are incorporated to quantitatively measure system performance under disaster situations. As chance constraints are intractable, we correspondingly propose conic and linear approximations. The reformulated model is solved within the outer approximation framework, where three acceleration techniques, i.e., the branch-and-cut algorithm, in–out algorithm, and Benders decomposition, are embedded to increase the computational efficacy. Through extensive numerical results and a case study, our proposed model is found to be superior to traditional scenario-based approaches. • We consider a location and sizing problem for disasters with chance constraints. • Efficiency, effectiveness, and equity are involved simultaneously. • Demand and supply uncertainties are handled by robust optimization. • Computation efficacy is accelerated by three outer approximation algorithms. • A case study under hurricane threat is conducted to illustrate the superiorities.