Greedy randomized adaptive search procedure for supply chain network resilience optimization considering risk diffusion

Abstract

Supply disruption risks occurring at individual nodes in a supply chain network can propagate to other nodes through the links in the network affecting the overall performance of the supply chain. In this context, mitigation strategies are commonly implemented to respond to these disruptions and improve the resilience of the supply chain. This study proposes a mathematical model to optimize the resilience of the supply chain network by selecting critical nodes and identifying the appropriate type of mitigation strategies to be implemented under a limited budget. The resilience of the supply chain network is evaluated by the degree of propagation of risks from individual nodes to the entire supply chain network, using a risk diffusion model where a disruption risk decreases exponentially as it progresses downstream in the supply chain. A greedy randomized adaptive search procedure (GRASP)-based solution approach was developed to solve the proposed formulation and its performance was compared to other solution procedures for different scenarios showing it to be the most effective and robust. A summary of the insights developed from the solutions obtained for the different scenarios is presented along with directions for future research.