Robust adaptive shear thickening fluid variable admittance control in image-based visual servo feature space

Abstract

To improve the compliance and robustness of robots during physical human–robot interaction (pHRI) and to enhance the success rate and stability of peg-in-hole experiments, a robust adaptive shear-thickening fluid (STF) variable admittance control in feature space is proposed in this article. Leveraging STF’s adaptability to contact forces and its energy dissipation properties during impacts, a robust adaptive virtual damping coefficient based on the maximum feature point of virtual velocity is designed. Additionally, an adaptive virtual stiffness coefficient that adjusts with contact force is developed to ensure rapid convergence of the manipulator when subjected to external impacts during servoing. The designed variable admittance controller ensures the manipulator’s compliance and robustness during pHRI. Furthermore, image-based visual servoing (IBVS) is coupled with admittance control in feature space to avoid the need for precise environmental modeling during tasks. Finally, the system’s stability is verified through the Lyapunov stability theorem. Experimental results demonstrate enhanced convergence and compliance during tasks, along with robustness and energy dissipation capabilities against external impacts.