Simply optimizing techniques to fine-tune UAVs

Dariusz Horla

Institute of Robotics and Machine Intelligence, Faculty of Control, Robotics and Electrical Engineering
Poznan University of Technology

Abstract

During the seminar, a novel auto-tuning approach for finding locally-best parameters of controllers on board of unmanned aerial vehicles (UAVs) will be presented. The controller tuning is performed fully autonomously during flight on the basis of predefined ranges of controller parameters. Required controller properties may be simply interpreted by a cost function, which is involved in the optimization process. For example, the sum of absolute values of the tracking error samples or performance indices, including weighted functions of control signal samples, can be penalized to achieve very precise position control, if required. The proposed method relies on a zero-order optimization search technique fitted into bootstrap sequences, enabling one to obtain a global minimizer for a unimodal cost function. The approach is characterized by an extremely low computational complexity and does not require any UAV dynamics model (just periodic measurements from basic on-board sensors) to obtain proper tuning of a controller. In addition to outlining the complete, yet simple, theoretical background of the method, an experimental verification in real-world outdoor conditions will be provided in relation to multiple experiments, applications, and results.

The seminar is mainly suited for researchers, Ph.D. students, and M.Sc.-level students.

Biosketch

DARIUSZ HORLA, Member, IEEE, received his M.Sc. degree from Poznan University of Technology in 2002, his Ph.D. degree in 2005, and his D.Sc. degree in 2013 from the same university. Since 2020, he has been an associate professor at PUT, and since 2023, a full professor. He was a visiting scholar at the University of Madrid, FCT Nova in Lisbon, Universidad de Cantabria, Universita di Catania, UNINA, and Czech Technical University in Prague. He is the (co)author of 60+ publications in peer-reviewed conferences, 30+ publications in impacted journals, and 40+ papers in other journals. He focuses on using control or optimization techniques, optimal, adaptive, or robust control methods, developing linear matrix inequality conditions for anti-windup compensation, and tuning problems. He was a member of the UPM-UPO-PUT team in the MBZIRC 2020 robotic competitions (Abu Dhabi) where they scored third place in the Grand Challenge. He took part in four H2020-cascaded projects with LAAS-CNRS, EPFL, Imperial College London, and Twente University.