IMU-Based Exoskeleton Control: Torso Movements and AI

Abstract

This study introduces a new control system for an upper limb exoskeleton, leveraging Inertial Measurement Unit (IMU) sensors placed on the user's trunk. The system employs two distinct control methodologies to enhance the exoskeleton's responsiveness and accuracy in assisting arm movements. The first method utilizes the torso's motion, integrating IMU data to calculate the arm's movement limits synchronously with the torso, ensuring the exoskeleton's movements are in harmony with the user's natural motion patterns. The second method adopts a more advanced approach, employing a neural network to predict the user's intended arm movement based on the torso's dynamics. This predictive model allows for a more intuitive interaction between the user and the exoskeleton, potentially improving the efficiency and satisfaction in its use. By comparing these methods, the paper aims to evaluate their effectiveness in providing a seamless and natural extension of the human body through the exoskeleton, offering insights into future developments for assistive technologies.