ABSTRACT
Drying can be seen as an operation where the objective is to reach some final food quality using for this purpose some optimally spent thermal energy. Because of its wide variety of products, drying technologies, production objectives, etc., the drying process presents some specific problems for automatic control. Nonlinearity is probably the main problem. We have shown that several interesting strategies are already used or in validation study. A linearization technique has proven, in several applications, to be efficient and simple. It can be used in most cases to increase the robustness of the controllers. Model-based predictive control appears to be one of the most promising techniques to treat nonlinear and multivariable problems as are found in drying. The model type is no longer a limitation for this method, so that it is interesting to combine it with a first-principle model.
