Vision-based finite-time prescribed performance control for uncooperative UAV target-tracking subject to field of view constraints

• Published in: ISA transactions Vol. 149; pp. 168 - 177
• Main Authors: Sun, Peng, Li, Siqi, Zhu, Bing, Zheng, Zewei, Zuo, Zongyu
• Format: Journal Article
• Language: English
• Published: United States Elsevier Ltd 01.06.2024
• Subjects: Finite-time control; Prescribed performance; UAV; Vision-based control; Finite-time control; UAV; Prescribed performance; Vision-based control
• DOI: 10.1016/j.isatra.2024.04.014
• ISSN: 0019-0578; 1879-2022

This paper presents a vision-based finite-time prescribed performance controller for unmanned aerial vehicle (UAV) tracking of uncooperative aerial targets. The relative states between UAV and target are estimated by an onboard monocular camera. The inability of visual measurements to accurately determine the initial state of the target renders conventional prescribed performance controllers ineffective in such situations. As a result, it becomes essential to address the problem of prescribed performance control under conditions of uncertain initial values By utilizing an auxiliary transforming function, an Asymmetric Barrier Lyapunov Function (ABLF) and a finite-time prescribed performance function, a robust adaptive controller based on backstepping framework is proposed to deal with state constraints under unknown initial tracking conditions. It is proved that, the closed-loop relative position is capable of reaching the prescribed performance bound before the preset transforming time and converging to the prescribed steady-state error before a finite setting time. Simulation examples are provided to illustrated the effectiveness of the proposed tracking algorithm. •An adaptive strategy is proposed to converge tracking error within predefined performance boundaries within finite time, despite initial value constraints.•A transformation function ensures that the transformed tracking error remains within prescribed performance bounds, and faster convergence can be achieved by adjusting its parameters.•The prescribed performance function is tailored to image boundaries, enabling effective visual tracking control of non-cooperative targets by UAVs.