Résumé
A semi-empirical nonlinear model of moisture content and wet-milling quality degradation in corn drying is established. The model uncertainty is expressed as a relative error level which includes a pre-specified percentage of experimental data. An optimal control strategy for batch drying is determined which maximizes the dryer throughput, while simultaneously satisfying two quality constraints on the final state: achieving specified moisture content and wet-milling quality levels. Taking into account the model uncertainty results in a conservative but less efficient control strategy. The performance - robustness compromise is discussed.