A number of algorithms have been developed -including enumeration of feasible production sequences, alternative task selection and the generation of alternative production lines- to determine the optimal sequence in which products and by-products should be produced and the times at which the various production operations for each product should be carried out to meet a given product demand pattern, taking into account the available equipment, storage costs, stopover penalties and other plant limitations.
Product interdependencies and utility requirements and constraints which limit the number of operations that can be carried out in parallel, are also considered. These algorithms have been integrated into a single, efficient computer package that, integrated in presently available low-cost workstations, is intended for routine use in industrial operational practice. A user-friendly menu-driven interface is also available.
The proposed methodology permits the solution to real, complex problems thanks to the use of a new, improved hybrid combination of heuristic and deterministic algorithms. Production planning in an already existing plant has also been successfully achieved for short and medium-term production using problem decomposition techniques and integer-linear programming with restrictions. The algorithmes have proved to be very fast when compared with existing deterministic methods. The industrial cases presented here demonstrate the practical use of the package under different plant production conditions.
