Robust Passivity-Based Dynamical Systems for Compliant Motion Adaptation

Abstract

Motivated by human compliant behaviors during interacting with unknown environments and how motions and impedance to are adapted skilfully complete a task, this article develops a motion planning scheme that is capable of generating a compliant trajectory online such that tracking desired contacting forces under a predefined motion task. First, an improved dynamical system (DS) is designed to generate an adaptive compliant scanning trajectory online from the original DS in terms of the contact forces and the desired scanning forces. Inspired by passivity analysis for the robot control system, a robust term is formulated to guarantee stability by considering the balance between environmental and robotic energy. Furthermore, we develop a state-constrained controller based on barrier Lyapunov function to track the compliant DS motion and to ensure safety during scanning for the patient. Finally, comparative simulations are conducted to validate the general compliant capability of the proposed framework. We also instantiate our methodology through a use case of liver ultrasound scanning to demonstrate the stable and dynamic force tracking performance.