Augmented Robust Model Predictive Control Design of a Rocket based on a Disturbance Observer

This paper presents a novel robust solution to the problem of model predictive control for a nonlinear, discrete-time rocket in the presence of finite disturbances. First, a mathematical model for the rocket in the state space was presented. Then, the basic equations of a typical six-Degree-of-Freedom airframe dynamics (6DoF) were separated as lateral and longitudinal dynamic equations. Linearization of these coupled dynamics was presented by using aerodynamic coefficients. Next, an augmented Model Predictive Control (MPC) was designed by using an observer to estimate states and disturbances, allowing the controller to reject disturbances. Application of the disturbance observer leads to the definition of a new state space and domain for MPC, which was considered in the present research. The predictive model control problem for the uncertain system was solved in finite time and online, and the decision variables were the initial states and disturbance estimations. Finally, the performance of the developed method was evaluated by simulation.