Good dynamic models are fundamental for the conception, operation and optimization of any complex system, and bioreactors are no exception to this rule. For operation of a given bioreactor, precise prediction of the reactor’s output is the main purpose of a model, and Data-Driven (black-box) models are most convenient for this job. Identification of DD models for bioreactors using classical model structures (NAR, NARMA, NARMAX, etc) and/or artificial neural networks (ANNs) is the aim of this MSc project.
The student will follow courses on Linear Systems, Nonlinear Systems and System Identification. Her/his work will include measuring data from real bioreactors; treating these data and then using them for the identification of nonlinear DD models; validating the models and comparing their performance; assessing the improvement in the efficiency of the bioreactor’s operation by the use of these models.
