A robust sliding mode observer for non-linear uncertain biochemical systems

A problem of state estimation for a certain class of non-linear uncertain systems has been addressed in this paper. In particular, a sliding mode observer has been derived to produce robust and stable estimates of the state variables. The stability and robustness of the proposed sliding mode observer have been investigated under parametric and unstructured uncertainty in the system dynamics. In order to ensure an unambiguous non-linear state (coordinates) transformation, the appropriate system model for the observer synthesis has been devised and analysed. The stability analysis of dynamics of estimation error has been carried out, based on the Lyapunov stability theory in relation to Lipschitz assumptions for non-linear functions. In order to validate the performance of the devised observer, it has been applied to the model of a continuous stirred tank reactor (bioreactor). The promising simulation results have been obtained and they demonstrate the high effectiveness of the devised approach.