A nonlinear high gain observer for an olive oil waste esterification in a continuous stirred tank reactor

Abstract

In this paper, a nonlinear high gain observer is developed to reconstruct the full state of an olive oil waste esterification process. This operation is taking place in a semi-continuous stirred tank reactor. The plant model, as most of the chemical processes, presents a high level of non linearities and interconnections between states. Most of the works developed around such model are based on a canonical form of the model or in single input single output structure. We show, in this work, that it is possible to synthesis a high gain based full observer for a multi input multi output system and to ensure good applicability of a such observer. The results are validated on real measurements of the plant. The second part is devoted to performances' comparison between our observer and the well used observer for such nonlinear systems, the Extended Kalman Filter. The comparison shows that the high gain observer presents better results in precision and application limitations.