Optimal operation of fed-batch bioreactors is an important practical issue. Since control actions are saturation-limited, the optimal control consists usually of both singular and bang-bang arcs. However, its realization requires good model knowledge and also the measurement of all state variables, requirements rarely satisfied in real applications. In this paper a method is proposed, for a class of bioreactors, to robustly optimize the operation when few measured variables are available and the model is uncertain. Such control law is justified and its properties analysed. The case of a bioreactor for treating toxicants in a waste water treatment plant is thoroughly studied, and experimental results in a lab-scale bioreactor are presented. The method can be also applied for other applications, as in biotechnology.
|Número de páginas||18|
|Publicación||International Journal of Robust and Nonlinear Control|
|Estado||Publicada - feb. 2006|