History-based purchase-inventory optimization model and global sensitivity analysis in iron and steel industry

This paper investigates a purchase-inventory optimization problem arising from the operation management practice in iron and steel industry. In the cost-oriented purchase problem, the demands during future planning periods are unknown, and we propose a new history-based model for optimizing the purchase and inventory in steel company. The model makes use of the rich historical demand information which can guide the purchase and inventory decision making. In addition, we use the real date from a domestic steel company and test the model, and the computational results show the model is effective. In addition, we apply the Morris method and test a series of combination of cost parameters, and then the results of global sensitivity analysis show that the purchase decisions are robust.