Algorithms for Numerical Modeling of Supply Chains in the Electrical Circular Economy

This study develops a multiobjective mixed-integer nonlinear programming framework for circular supply chains in the electrical and electronic equipment (EEE) sector under operational disruptions. The formulation integrates strategic network decisions with dynamic material flows, congestion-dependent transit times, delay-related quality degradation, nonlinear recovery yields, inventory evolution, service backlogs, and operational risk measured through conditional value-at-risk. An illustrative 2024 planning instance comprising 126 feasible policies was evaluated through exhaustive enumeration, which identified an exact reference front of 78 nondominated policies. MOEA/D, NSGA-II, and uniform random sampling were subsequently compared over 30 independent runs using a common budget of 560 objective-function evaluations. Uniform random sampling achieved the highest relative hypervolume (99.91%), the lowest IGD+ (0.000313), and the greatest exact-front coverage (98.85%). NSGA-II outperformed MOEA/D, attaining respective hypervolume values of 98.99% and 97.59% and exact-front coverage rates of 84.83% and 70.90%. Because the evaluation budget exceeded the finite decision-space size by more than four times, these differences were influenced substantially by duplicate candidate evaluations and should not be interpreted as evidence of the general superiority of random sampling. The results instead highlight the importance of matching the solution method to the size and structure of the decision space. The framework provides a transparent basis for examining trade-offs among penalized cost, recovered output, and disruption-related operational risk.