Résumé
The identification of model parameters for food processes is both an experimental and numerical challenge. The optimal design of experiments aims to enhance the experimental strategy necessary to achieve the best identification results. Here, we discuss the trade-offs between improvements of pilot, experimental efforts, computational efforts, and the precision of the parameters identified. To identify four parameters related to heat and mass transfer for a cereal drying model, it was found that the best return on investment was to implement high speed and precise in-line measurements. This investment enabled an estimated precision better than 20% on parameter estimates based on only two experiments. We also discuss the generalization of this method to other domains.