An intelligent digital twin approach for optimizing multi-channel supply chains in uncertain environments
Hamed Nozari, Zornitsa Yordanova
Abstract
This study develops an innovative framework for the multi-objective optimization of multi-channel supply chains under uncertainty. The framework integrates the D-number model, an extension of Dempster–Shafer Theory (DST) that enables the representation of incomplete and non-exclusive information, and employs a cognitive digital twin (CDT) as a platform for analysis and decision-making. The proposed mathematical model captures the complexity of the supply chain environment by addressing conflicting objectives such as minimizing total cost and delivery delays while enhancing resilience. Small-scale instances are solved using the General Algebraic Modeling System (GAMS), while two meta-heuristic algorithms, the Non-dominated Sorting Genetic Algorithm II (NSGA-II) and the Colonial Competitive Algorithm (CCA), are applied to larger problems. Sensitivity analysis, Pareto front comparisons, and a scalability study demonstrate that the framework remains robust across varying conditions and offers valuable managerial insights. The results support improved decision-making and enhanced supply chain resilience. • Integrates D-number belief model with cognitive digital twin for multi-objective SCM. • Balances cost, delivery delays, and resilience in multi-channel supply chains. • NSGA-II outperforms CCA in solution quality and Pareto front diversity. • Meta-heuristics scale efficiently—GAMS only viable in small problem instances. • Increased fleet capacity under uncertainty significantly improves performance.