Designing an Agile, flexible and resilient disaster supply chain network using a hybrid group decision-making robust optimization framework

Abstract

Natural disasters are a part of human history worldwide. To mitigate after-disaster damages, it is imperative to create an effective disaster supply chain. This research proposes a stochastic multi-objective mixed-integer linear programming model for an agile, flexible disaster supply chain network. Selecting domestic first-level suppliers and foreign second-level suppliers is a complicated process. We propose a novel group decision-making (GDM) framework that utilizes k-means clustering, the Borda count method, a consensus-reaching process, and the best-worst method. This framework results in the selection of qualified suppliers at the first and second levels. The GDM process utilizes agility indicators to reduce the delivery time of commodities. A robust probabilistic programming approach is used because of the uncertainty in some critical parameters in the crisis. To solve the large-scale problems, three well-known metaheuristic algorithms are employed. Metaheuristics parameters are then set to their optimum levels using the Taguchi method. Validations of the model and the solution methodologies used are based on a case study with twelve datasets. The obtained results indicate the applicability of both our group decision-making framework and the solution methodologies employed to solve the robust optimization model.