Computationally efficient receding horizon trajectory tracking control for a tubular reactor example



Flatness-based feedforward control and receding horizon trajectory tracking control can be combined within a two-degrees-of-freedom (2dof) control scheme to obtain robust trajectory tracking for nonlinear systems. The key to a computationally efficient solution of the optimization problem underlying the receding horizon control is the linearization of the time-variant tracking error dynamics such that the nonlinear optimal control problem may be replaced by a linear-quadratic formulation. The approach is demonstrated for setpoint transitions of a nonlinear boundary-controlled parabolic partial differential equation (PDE), which is semi-discretized for the control design using finite differences on the spatial derivatives. (© 2012 Wiley-VCH Verlag GmbH & Co. KGaA, Weinheim)