Ranjith M

The flatness of a six-degree-of-freedom (6DoF) aircraft model with conventional control surfaces - aileron, flap, rudder and elevator, along with thrust vectoring ability is established in this work. Trajectory optimisation of an aircraft can be cast as an inverse problem where the solution for control inputs that yield desired trajectories for certain states is sought. The solution to the inverse problems for certain systems is made tractable when they exhibit differential flatness. Flatness-based trajectory optimisation has a significant advantage over an equivalent collocation-based method in terms of computational efficiency and viability for real-time implementation. An application for the flatness of 6DoF aircraft is shown in the trajectory optimisation for dynamic soaring, and its connection with an equivalent 3DoF flatness-based implementation is also brought out. The results are compared with that from a collocation-based approach.

Trajectory tracking using linear model predictive control (LMPC) ensures optimal tracking of the system trajectory subject to constraints imposed by system dynamics and actuator limitations. Even though the method is attractive for most applications, solving an optimization problem tends to be a computationally intensive task. In order to alleviate the computational cost associated with LMPC, a flatness based model predictive control (FMPC) algorithm is proposed and simulated for a 6DoF UAV. The reduction in the dimension of the optimization problem due to flatness leads to a significant increase in computational efficiency. In order to show the computational advantages of the FMPC algorithm, it is compared with a standard LMPC algorithm.