Tiwari , Shivendra N and Padhi, Radhakant (2013) DR-SNAC Aided Dynamic Inversion Controller for Robust Trajectory Tracking of a Micro-Quadrotor. In: AIAA Guidance, Navigation, and Control (GNC) Conference, Guidance, Navigation, and Control and Co-located Conferences, (AIAA 2013-4533) , August 19-22, 2013, Boston, Massachusetts.
AIAA_Gui_Nav_Con_Con_2013 .pdf - Published Version
Restricted to Registered users only
Download (943Kb) | Request a copy
A dynamically reoptimized single network adaptive critic (SNAC) aided dynamic inversion controller is presented in this paper, where an offline trained SNAC op-timal controller is modified with the help of another neural network trained online so as to dynamically re-optimize it for the real plant. This is done by updating the plant model in real-time with the help of yet another neural network trained online. The resulting optimal controller is then mapped to a dynamic inversion controller structure online. As a result, it brings in two major advantages, namely (i) a stabi-lizing closed form expression of the controller and (ii) easy scalability to command tracking application even without apriori knowledge of the reference command. Ef-fectiveness of the proposed controller is demonstrated by applying it for attitude control and trajectory tracking of a micro-quadrotor, which is a challenging prob-lem due to its small size, negligible aerodynamic damping as well as strong coupling in pitch-yaw-roll channels. It has been demonstrated from six degree-of-freedom simulations that the controller is capable of achieving the intended objective and has good robustness as well.
|Item Type:||Conference Proceedings|
|Additional Information:||Copy right for this article belongs to the AIAA Guidance, Navigation, and Control (GNC) Conference, Guidance, Navigation, and Control and Co-located Conferences, (AIAA 2013-4533)|
|Department/Centre:||Division of Mechanical Sciences > Aerospace Engineering (Formerly, Aeronautical Engineering)|
|Date Deposited:||25 Aug 2016 10:13|
|Last Modified:||25 Aug 2016 10:13|
Actions (login required)