Distributed impedance control of coordinated dissimilar upper-limb exoskeleton arms

Abstract

Upper-limb exoskeleton (ULE) arms have emerged as promising wearable robotic devices for assisting humans in accomplishing tasks by distributing heavy loads. However, the complex structures, actuators, and power transmissions inherent in ULE arms introduce challenges in their control system. Additionally, the coordination of dissimilar ULE arm systems, stemming from differences in vendor structures and biomechanical variations among users, further complicates the control approach. In response to these challenges, this paper presents a new distributed framework for adaptive impedance-based Virtual Decomposition Control (VDC) tailored to ULE arms. It attains enhanced performance, robustness, and coordination by integrating prediction capabilities and ancillary control laws, which are significant for dissimilar ULE arms when manipulating a common object. Stability analysis is conducted using the input-to-state stability approach. We validate our approach through simulations with six coordinated ULE arms and extensive experiments with two commercial ULE arms, each with seven degrees of freedom, in four scenarios, including obstacle interaction. Comparative analysis against a state-of-the-art adaptive VDC method demonstrates the superiority of our approach. The proposed method enhances performance, robustness, and coordinated control, enabling safe and efficient manipulation across various scenarios.