Application of Model Predictive Control to Position and Height Limitation of a Quadrotor Unmanned Aerial Vehicle

This paper discusses a position and height limitation control for a quadrotor UAV (Unmanned Aerial Vehicle) using Model Predictive Control (MPC) approach. Nonlinear dynamics of the quadrotor is discussed first, and decoupled linearized dynamics is obtained. For the implementation of MPC, extended state vector of vehicle is generated, and augmented linear dynamics is constructed. The MPC in this paper utilizes a set of Laguerre function as basis to approximate the future movement of modeled vehicle. Position/height constraints and vehicle actuator characteristics enter the dynamics as linearized inequalities, which could be solved on-line via a recursive optimization approach. While validations based on experimental tests will be conducted in future, currently simulations have been completed. Based on the simulation results, when state of the vehicle is laid within the permissible bound, it retains the same dynamics of original vehicle. However, if predicted response exceeds the limits, however, MPC will take effect and restrict associate vehicle states. The discussed MPC framework in this paper is considered to be applicable.