Hierarchical mathematical programming for operational planning in a process industry

Scheduling techniques for discrete manufacturing industries do not fit in many process industries. A technique that often is used in process industries is mathematical programming. However, mathematical programming is used mostly for strategic or tactical planning. This article describes a problem at the operational planning level in a process industry. The (large) problem is decomposed into smaller problems, which can be solved separately with use of mathematical programming. The resulting models are less complex and require less calculation time than a large model. The decomposition is based on the uncertainties for raw materials and customer orders that exist for future time periods. In this way a Decision Support System is developed which has an acceptable calculation time.