An extended version of adaptive large neighborhood search for a relief commodities distribution network design under uncertainty

Abstract

Natural and technology-induced disasters have posed significant threats to human life all around the world and caused many damages and losses so far. The current study addresses a location-routing problem to make an efficient and timely distribution plan in response to a possible earthquake. This problem considers uncertainty in such parameters as demand, access to routes, travel time, and the number of available vehicles. To deal with these uncertainties, stochastic programming (SP) is performed while the objective function is to minimize the time of carrying relief commodities (RCs) to affected areas. The problem is coded in the CPLEX solver to obtain optimal solutions to small-scale problems, and an adaptive large neighborhood search (ALNS) is proposed to solve mixed-integer linear formulas for large-scale problems. To validate the formulation and evaluate the performance of the proposed ALNS, several types of tests are devised. To shows the efficiency of the proposed ALNS, two other metaheuristic algorithms, the Genetic algorithm (GA) and simulated annealing algorithm (SA), are used as well. The results of the calculations suggest the satisfactory performance of the suggested algorithm and the effectiveness of the model for the desirable delivery of humanitarian aids to affected areas.