Two-Stage Robust Optimization for CircularDisassembly Line Balancing with Budgeted TimeUncertainty

Abstract

The proliferation of end-of-life(EOL) electronic products necessitates efficient disassembly recycling, yet fluctuations in task times caused by equipment wear and operator variability compromise deterministic line-balancing schedules. This paper addresses the circular disassembly line balancing problem under budgeted task-time uncertainty, where workstations are arranged radially to improve space utilization and component feeding flexibility. A profit-maximizing two-stage robust optimization model is established, in which a cardinality-constrained uncertainty set controls the conservatism level against simultaneous worst-case task durations. The resulting max-max-min formulation is solved by a tailored L-shaped decomposition that separates first-stage task assignment and workstation activation from second-stage worst-case delay evaluation via strong duality. Extensive experiments on small, medium, and large-scale instances demonstrate rapid algorithmic convergence and quantify the profit-robustness trade-off as the uncertainty budget increases.