Model predictive versus linear quadratic control for the tracking problem of automatic river navigation

Within this paper the tracking problem of automatic river navigation is presented. To realize “track-keeping” control on rivers the representation of a desired track by splines is introduced. First, a linear quadratic regulator combined with feedforward control is used to solve the control task. Second, nonlinear model predictive control is realized. Besides the “classical” model predictive control a multilayer control structure involving prediction is presented. Since a practical application is considered also real-time demands are addressed.