Exploiting Non-Linear Model Predictive Control in Contact-based Aerial Physical Interaction

Finished: 2021-10-29

MSc assignment

This MSc thesis assignment concerns the integration of an existing Nonlinear Model Predictive Control (NMPC) framework with interaction control of fully actuated UAV to be used in aerial interaction with 3D surfaces. As the starting point, the idea is to extend an existing NMPC work, currently implemented in a Matlab-Simulink framework [1], where a controller was designed to perform trajectory tracking and local planning for a generic multi-robot platform. 

The main objective of this Master Thesis is to extend the aforementioned work with sensor-less wrench tracking procedure and accomplish contact-based interaction task [2], either with 3D surfaces or for collaborative tasks with human operators. This can be achieved by adding an outer loop controller and modifying the cost function based on the new constraints. There might be the possibility to benefit from other methodologies such as coordinate-free dynamics [3] to increase precision on trajectory tracking of the end-effector. 

The objective is to accomplish interaction tasks with the aerial robot endowed with a static tool, first, and possibly extend the method to also control an onboard 3 DoF arm manipulator mounted. The control of the floating platform and the arm can be also done in a decoupled way if the model of the entire manipulator cannot be developed for a matter of time. In any case, the work should be developed in an incremental way. The controller should be interfaced with the software architecture used at the RAM group (which uses an extension of ROS called GenoM), in particular with the gazebo simulator. Final validation of the controller is aimed to be performed experimentally on an available fully-actuated multi-rotor UAV developed at RAM. The use of the RAM architecture (based on GenoM) is a requirement for software implementation. 

This MSc thesis will take place within the scope of the H2020 Aerial-Core1 European project and the RAAK PRO Mars4Earth project2. It should provide relevant results to both projects within the scope of aerial interaction for applications such as the inspection and maintenance of large linear infrastructures and interactive tasks with human operators. 

References:

[1] Bicego, D., Mazzetto, J., Carli, R., Farina M., and Franchi A. "Nonlinear Model Predictive Control with Enhanced Actuator Model for Multi-Rotor Aerial Vehicles with Generic Designs’’. J Intell Robot Syst, 2020, 100, 1213–1247 (2020). https://doi.org/10.1007/s10846-020-01250-9  

[2] Ryll M., Muscio G., Pierri F., Cataldi E., Antonelli G., Caccavale F., Bicego D.,and Franchi A., "6D interaction control with aerial robots: The flying end-effector paradigm’’. IJRR, 2019, 38 (9), pp.1045-1062. https://doi.org/10.1177/0278364919856694  

[3] J. Welde and V. Kumar, "Coordinate-Free Dynamics and Differential Flatness of a Class of 6DOF Aerial Manipulators," 2020 IEEE International Conference on Robotics and Automation (ICRA), Paris, France, 2020, pp. 4307-4313, doi: http://10.0.4.85/ICRA40945.2020.9196705

[4] D. Lunni, A. Santamaria-Navarro, R. Rossi, P. Rocco, L. Bascetta and J. Andrade-Cetto, "Nonlinear model predictive control for aerial manipulation," 2017 International Conference on Unmanned Aircraft Systems (ICUAS), Miami, FL, USA, 2017, pp. 87-93, doi: 10.1109/ICUAS.2017.7991347. 

[5] G. Darivianakis, K. Alexis, M. Burri and R. Siegwart, "Hybrid predictive control for aerial robotic physical interaction towards inspection operations," 2014 IEEE International Conference on Robotics and Automation (ICRA), Hong Kong, China, 2014, pp. 53-58, doi: 10.1109/ICRA.2014.6906589. 

[6] B. Yüksel, C. Secchi, H. H. Bülthoff and A. Franchi, "A nonlinear force observer for quadrotors and application to physical interactive tasks," 2014 IEEE/ASME International Conference on Advanced Intelligent Mechatronics, Besacon, France, 2014, pp. 433-440, doi: 10.1109/AIM.2014.6878116. 

[7] L. Peric, M. Brunner, K. Bodie, M. Tognon, R. Siegwart, "Direct Force and Pose NMPC with Multiple Interaction Modes for Aerial Push-and-Slide Operations", 2021 IEEE International Conference on Robotics and Automation (ICRA), Xi'an, China, 2021, https://www.research-collection.ethz.ch/bitstream/handle/20.500.11850/476947/root.pdf?sequence=1&isAllowed=y