This MSc thesis assignment concerns the study, design, and implementation of a control-law for multi-rotor UAVs based on optimal control in a Nonlinear Model Predictive Control (NMPC) fashion, which has to be tailored for preliminary aerial physical interaction applications, in the view of future applications involving cooperation with a human operator. As a starting point, the idea is to extend an existing work, currently implemented in a Matlab-Simulink framework (cf. [1]) where a controller was designed to perform trajectory tracking and local-planning for a generic multi-rotor platform.
The leitmotif of this MSc assignment is to extend the aforementioned previous work to make the controller aware of the external wrench produced by the external environment and, eventually, accomplish preliminary interaction tasks which involve the tracking of a desired wrench on top of pure trajectory tracking, to achieve mixed motion/wrench control. This will be made possible by integrating a wrench observer in the loop and by properly modifying the control law to account for the external action and the required physical interaction task.
The MSc thesis will take place in the scope of Aerial-Core, cf. https://aerial-core.eu/, a European H2020 Project with the goal of developing an integrated aerial cognitive robotic system with unprecedented capabilities on the operational range and safety in the interaction with human co-workers for applications such as the inspection and maintenance of large linear infrastructures.
[1] D. Bicego and J. Mazzetto and M. Farina and R. Carli and A. Franchi, "Nonlinear Model Predictive Control with Enhanced Actuator Model for Multi-Rotor Aerial Vehicles with Generic Designs", in Springer Journal of Intelligent & Robotic Systems, 2020, DOI: 10.1007/s10846-020-01250-9, https://arxiv.org/abs/1911.08183