Title: A hybrid approach for modelling dynamic flows and systemic risks in supply chains

Authors: Ke Sun; James T. Luxhøj

Addresses: Department of Industrial and Systems Engineering, Rutgers University, 96 Frelinghuysen Road Piscataway, 08854–8018, New Jersey, USA ' Department of Industrial and Systems Engineering, Rutgers University, 96 Frelinghuysen Road Piscataway, 08854–8018, New Jersey, USA

Abstract: The uncertainty of supply chain operations is becoming more complex with the growth of globalised business collaborations. Temporal business environment fluctuations cause disruptions that require quantitative risk analysis tools to understand and mitigate supply chain risks. This paper presents the formulation and application of the dynamic flow Bayesian networks (DFBNs) and the optimised dynamic flow Bayesian networks (ODFBNs) for a three-stage supply chain. DFBNs are created by integrating dynamic Bayesian networks (DBNs) and system dynamics (SD) to demonstrate the feedback flows of a supply chain with systemic risks considered. ODFBNs that incorporate mathematical optimisation with the original DFBNs are also developed to generate an optimisation-enhanced risk-influenced dynamic flow analysis. DFBN and ODFBN provide supply chain practitioners with a more effective reference for their business strategy. Comparison between the ODFBN and the DFBN is illustrated with a discussion of preliminary modelling results.

Keywords: supply chain; risk analysis; dynamic Bayesian networks; DBNs; system dynamics; multi-objective optimisation; hybrid model.

DOI: 10.1504/IJISE.2019.104276

International Journal of Industrial and Systems Engineering, 2019 Vol.33 No.4, pp.513 - 541

Received: 04 Mar 2017
Accepted: 30 Mar 2018
Published online: 24 Dec 2019 *

Full-text access for editors Full-text access for subscribers Purchase this article Comment on this article