A Stochastic Inventory Approach for Optimal Planning of Flexible Process Networks under Uncertainty

Due to global competition and dynamic market environment, a major challenge in enterprisewide optimization [1-3] has become to simultaneously optimize the production planning and inventory control decisions for chemical complexes under demand and supply uncertainty [4, 5]. Because chemical complexes are usually composed of many interconnected processes and various chemicals, it is an important but non-trivial task to integrate the production and inventory planning decisions with the uncertain supply and demand variations, and coordinate the activities of purchase, production, inventory and sale to minimize the total cost. The challenges arise not only from the modeling part about how to quantify the internal demand uncertainty and its propagation, but also from the computational perspective about how to effectively solve the resulting optimization problem that leads to a large-scale nonconvex mixed-integer nonlinear program (MINLP).