Vision-Guided Contact Force Control of an Omnidirectional Unmanned Aerial Vehicle

Recently, unmanned aerial vehicles have gained significant attention due to their high mobility when interacting with the surrounding environment. This paper presents an image-based admittance control strategy designed for regulating contact force. The proposed strategy adopts a two-loop structure. The outer loop utilizes a visual admittance filter to dynamically update reference visual features. This adaptation is based on force errors projected into the image space. The inner loop consists of a thrust controller for visual servoing that tracks the reference visual features. The visual tracking errors are incorporated into the admittance filter. This consideration is particularly pertinent during the contact phase, where perfect tracking of reference visual features from the admittance filter is unattainable. Moreover, to address the challenge posed by force data oscillations and delays during the interaction, we incorporate force saturation within the inner dynamic controller. This significantly enhances system performance. The stability of the system is proved using the Lyapunov method, treating the two loops as a unified framework. Finally, experiments using an omnidirectional unmanned aerial vehicle are conducted to demonstrate the effectiveness of the proposed method.