Rocket Landing Guidance with Linearized Model Predictive Control via Second-order Cone Programming

This paper aims to suggest a novel powered landing guidance for reusable launch vehicle to handle the disturbances and model uncertainties. To achieve this, we introduce a real-time optimization algorithm based on convex programming, complemented by an additional model predictive tracking guidance approach. In the tracking guidance, we use a linearized model to eliminate the need for successive convex programming and introduce new second-order cone constraints to address thrust and tilt angle limitations. Since it is not possible to use the lossless convexification in tracking problem, the proposed constraint can be helpful. Furthermore, since the tracking problem has a form of second-order cone programming, it can be guaranteed that this algorithm converges to global minimum in polynomial time, which is greatly attractive for guidance algorithm. The proposed algorithm is analyzed through a 3-DOF simulation, and its robustness is validated via Monte Carlo test.