Output feedback predictive control for hydrographic process

This paper introduces a new output feedback control scheme, which combine nonlinear generalized predictive control (NGPC) with a high gain observer (HGO). We consider the input affine nonlinear MIMO class of systems, that include the model of our hydrographic process. Indeed, from the liquid levels measurements in both bottoms tanks, we reconstruct the levels of both upper ones, and we use them in the state feedback control loop. The predictive control law synthesized is an explicit continuous solution of optimization problem, it arise from receding horizon index minimization. As it is known, to deal with the separation principle requirement, we need to lead the system in the new state coordinate and to impose control constraints. This last can inherently taking into account by the predictive controller, when we reformulate control law through the on-line solution of quadratic programming (QP) problem.