Safety-Critical Control Allocation for UAV Visibility-Preserving Surveillance

Abstract

Unmanned aerial vehicles (UAVs) are increasingly employed in surveillance tasks, where a key challenge lies in ensuring flight safety while continuously maintaining visibility of the target. Specifically, both collision avoidance and visibility-preserving constraints must be satisfied. Traditional methods often struggle to balance conservatism and computational efficiency. In addition, constraints are typically enforced at the planning or control-law level, which reduces design flexibility. This seminar introduces a safety-critical control allocation framework based on control barrier functions (CBFs), in which safety constraints are directly handled at the actuator allocation layer. By exploiting the geometric relationship between constraints and vehicle motion, a Conic Blind-Spot Control Barrier Function is proposed, enabling visibility-preserving surveillance with smoother avoidance trajectories and smaller control corrections.

Speaker

Dr Yang Haoyang is a Postdoctoral Fellow in the Department of Aeronautical and Aviation Engineering at The Hong Kong Polytechnic University. He received the BEng degree in Measurement and Control Technique and Instruments from Harbin Institute of Technology, Harbin, China, in 2017 and PhD degree in Navigation Guidance and Control from Beihang University (Beijing University of Aeronautics and Astronautics), Beijing, China in 2023. Before joining the PolyU, he was a Research Fellow at the University of Warwick, Coventry, UK. His research interests include safety-critical control, intelligent control, and their applications to spacecrafts and unmanned aerial vehicles. He has published in various top peer-reviewed journals, including IEEE Transactions on Aerospace and Electronic Systems, IEEE Transactions on Industrial Electronics, and IEEE/ASME Transactions on Mechatronics.