Nonlinear disturbance observer-based model predictive control for nonlinear system with unknown disturbances

This article presents a solution for accurate tracking control of disturbed nonlinear systems via an integrated framework that merges nonlinear model predictive control (MPC) with a nonlinear disturbance observer (NDO). First, a NDO is designed to compensate unknown disturbances and improve the model predict accuracy of MPC. Furthermore, based on the Taylor expansion, the proposed cost function of tracking errors is conducted, and the control input is given in explicit form, which also can reduce the computational load. Sufficient conditions are derived to ensure both tracking and anti-disturbance performance. Finally, comprehensive analyses of the simulations demonstrate the effectiveness of control performance, and verify the feasibility of potential applications on nonholonomic vehicles.