Aggregate production planning considering implementation error: A robust optimization approach using bi-level particle swarm optimization

Aggregate production planning (APP) is an important decision-making process for maintaining the effectiveness and responsiveness of manufacturing and supply chain systems. In this paper, we propose a robust optimization approach to the APP problem under implementation errors related to hiring and layoff. The APP problem with hiring and layoff uncertainty contain many equality constraints and integer uncertainty sets, a traditional robust counterpart might provide a conservative solution. To overcome this limitation, we develop a bi-level particle swarm optimization model to find an optimal solution that is always feasible and robust to unexpected variations in the workforce from hiring and layoff uncertainty. Experimental studies demonstrate that the proposed model outperforms a deterministic model and a traditional robust counterpart and in terms of the average cost and the cost of the worst-case scenario. The proposed model is also better than a conventional robust counterpart when the realized value is larger than the expected uncertainty, and suppresses the occurrence of a product shortage in supply.