Schedule of automotive manufacturing considering man-hour budget
Description
his dataset supports a study on human-robot collaborative scheduling in automotive Body-in-White (BIW) welding islands under the Industry 5.0 paradigm. The study formulates the scheduling problem as a multi-mode resource-constrained project scheduling problem (MRCPSP), considering precedence constraints, renewable resource limitations, alternative execution modes, and strict non-renewable man-hour budgets. The dataset includes benchmark scheduling instances, industrial BIW welding task data involving 43 tasks and 8 resource types, and computational results used to evaluate the proposed tri-stage adaptive simulated annealing (TS-ASA) algorithm. The TS-ASA framework integrates dynamic neighborhood operators, an adaptive cooling schedule, and a restart mechanism to improve exploration, convergence stability, and budget-constrained solution quality. Comparative experiments include ablation studies, hyperparameter sensitivity analyses, comparisons with GA, DPSO, EDA and rule-based FIFO-GM baselines, as well as validation against Gurobi on standard MRCPSP benchmark instances. The results demonstrate that TS-ASA achieves robust scheduling performance and reduces the industrial case cycle time by 13.54% compared with the rule-based baseline, indicating its potential value for flexible, high-efficiency manufacturing systems
Files
Institutions
- Changchun University of TechnologyJilin, Changchun