In this research, a force tracking smoothing adaptive admittance controller is proposed that grants precise contact forces (performance necessary for many critical interaction tasks such as polishing) for unknown interaction environments (e.g., leather or thin and soft materials). First, an online indirect adaptive update strategy is proposed for generating the reference trajectory required by the desired tracking force, considering the uncertainty of the interaction. The sensor noise amplitude is environment dynamics and the necessity condition for traditional admittance controller to achieve ideal steady-state force tracking. Then, a pre-PD controller is introduced to increase the parameter convergence rate while ensuring the steady-state force tracking accuracy and enhancing the robustness of the system. The robustness boundary is also analyzed to provide assurance for the stability of the system. Finally, we verify the effectiveness of the proposed method in simulations. Simultaneously, an experiment is conducted on the AUBO-i5 serial collaborative robot, and the experimental results proved the excellent comprehensive performance of the control framework.

In recent years, autonomous control based on contact inspections in unknown environments is a new hot and difficult point in robotics research. This paper presents a new control law for unmanned aerial manipulator (UAM) to perform contact inspection tasks on vertical surfaces. The selected circular image feature decouples the position and attitude of UAM, so an image-based impedance control is proposed to control the position and track the contact force. The developed controller uses geometric methods to control the attitude. In addition, the designed aerial manipulator decouples the roll and pitch of the UAV from the UAV, which improves the system’s stability. Experiments have been carried out to demonstrate the feasibility of this method.

In this paper, motion synchronization is investigated for human–robot collaboration such that the robot is able to “actively” follow its human partner. Force tracking is achieved with the proposed method under the impedance control framework, subject to uncertain human limb dynamics. Adaptive control is developed to deal with point-to-point movement, and learning control and neural networks control are developed to generate periodic and arbitrary continuous trajectories, respectively. Stability and tracking performance of the closed-loop system are discussed through rigorous analysis. The validity of the proposed method is verified through simulation and experiment studies.