Research on the robustness of interdependent supply networks with tunable parameters

Abstract

A supply network is considered robust when it can maintain operations and connectedness despite disruptions, and building robust supply networks is one of the challenges in supply chain management. Typically, a supply network is an interdependent supply network (ISN) which consists of an undirected cyber-layer network and a directed physical-layer network. The cascading failure in an ISN can be caused by over-load and loss-dependency failures simultaneously, called hybrid cascading failure (HCF), which is limitedly studied in previous studies. In addition, ISNs are usually characterized by some tunable parameters such as the number of nodes, the number of inter-links, interconnecting patterns, and scaling exponents. However, a generic model that can generate ISNs with all of the aforementioned parameters is not proposed, and hence how these parameters affect the robustness of ISNs when considering HCF is still not clear. Therefore, in this paper, we first propose an ISN model to generate ISNs with the aforementioned tunable parameters, and then study how these parameters affect the robustness of the generated ISNs against random and targeted disruptions when considering HCF. The simulation results mainly show that: (i) an ISN is very vulnerable to disruptions in most cases, and removing a small number of nodes can make it collapse; (ii) the behavior of an ISN is very complex when considering HCF; and (iii) the robustness of an ISN can undergo first- and second-order phase transitions under proper parameters. Our work may be helpful for developing HCF-mitigating strategies and building robust supply networks.