In this doctoral project, we are investigating the scheduling of process engineering multipurpose plants with discontinuous process control. The objective of the planning is to determine a production schedule with minimum cycle time for the production of given demand quantities. Existing solution approaches from the literature are usually based on the formulation of large mixed-integer linear programs that are solved using standard software. However, due to their complexity, problems of practical size cannot be solved with these approaches in a reasonable computation time. We therefore present a heuristic solution method based on the decomposition of the overall problem into a set and a scheduling problem.

Most scientific work on plant layout planning assumes static replenishment planning and deterministic planning data to solve the problem described. In reality, however, the production process takes place in a dynamic environment, so that events often occur during the implementation of schedules that were not taken into account in purely predictive planning. Examples of such events are the failure of operating resources, the extension of process times, losses in yields or the arrival of rush orders.

We therefore present a two-stage procedure in which we first design a schedule based on deterministic planning data that has the shortest possible cycle time. We then extract sequence and material flow relationships between the individual process step executions from the resulting predictive schedule, on the basis of which we control the production sequence. If disruptions occur in the course of production, we carry out revising planning, which either aims to minimize the cycle time of the new plan or to design a new plan that differs as little as possible in its structure from the original plan.