56 Optimisation of flour milling processes using the simulated annealing technique.

W. G. OWENS (1,2) and C. Webb (1). (1) Satake Centre for Grain Process Engineering, Department of Chemical Engineering, UMIST, P.O. Box 88, Manchester, M60 1QD, UK; (2) Odlum Group Ltd., Alexandra Road, Dublin 1, Ireland.

Optimisation of flour milling processes has generally concentrated on the economic performance of the business or of individual unit operations. The optimisation of milling processes, at operational level, using computer based techniques has not previously been attempted. Reasons include the complexity of the process in terms of the number of interdependent variables, the lack of mathematical models and demanding product criteria. This paper outlines a technique, known as simulated annealing, that can be applied to this problem. It can handle large numbers of variables, complex mathematical models, data sets instead of models and any number of constraints. Other strong features are that the algorithm is not influenced by the existence of local optimal solutions and the method systematically pursues a global optimal solution to a problem. The stepwise numerical basis of the technique means that it can be easily programmed and expanded to encompass any process requirements. A model of the break system in a flour mill has been optimised using this technique. The optimisation presents practical recommendations for mill settings that will maximise extraction within given constraints. This type of optimisation points the way towards significant process benefits encompassing key milling performance criteria.