This study aims to address the limitation in scheduling efficiency caused by subjective skill evaluations in apparel production lines by proposing a multi-objective worker scheduling optimization method based on the Analytic Hierarchy Process (AHP) and an improved Non-dominated Sorting Genetic Algorithm II (NSGA-II).
An evaluation system incorporating multidimensional skill indicators is first established using the AHP to determine factor weights, and workers' skill maturity is quantitatively assessed through normalization based on actual production data. A greedy initialization strategy and an adaptive mutation mechanism are then integrated into the NSGA-II framework to enhance convergence performance and solution space exploration, thereby improving its efficiency under practical scheduling constraints. Finally, optimization is conducted with skill maturity and production fluctuation minimization as objectives, yielding a set of near-optimal scheduling solutions.
Integrating AHP-based skill evaluation with an improved NSGA-II algorithm significantly enhances apparel production line worker scheduling, with the resulting solutions from case studies increasing workforce utilization by 11.5–15.3% and production output by 12.9–25.4%, while achieving faster convergence and higher solution quality.
The primary originality of this study lies in integrating AHP-based skill evaluation with production data, which is harnessed by an improved NSGA-II algorithm to efficiently tackle dynamic workforce allocation challenges, thereby significantly improving workforce allocation, computational efficiency, and solution quality.
