A hybrid genetic algorithm/fuzzy logic approach to manufacturing process control

Genetic algorithms and fuzzy logic have been proven to be effective in a variety of control applications. We present an approach for automatically determining the control parameters of a machine used in a manufacturing environment for producing cookies and other pastry. The proposed solution is based on augmenting a genetic algorithm with elements of fuzzy logic, since either of them alone could not produce satisfactory results. The feasibility of the approach is demonstrated by presenting experimental results obtained in a series of test productions. The tests show that high quality cookies can be produced in a very short time.