Reinforcement Learning and Control of a Lower Extremity Exoskeleton for Squat Assistance
Authors:
S. Luo,
G. Androwis,
S. Adamovich,
H. Su,
X. Zhou
Abstract:
A significant challenge for the control of a robotic lower extremity rehabilitation exoskeleton is to ensure stability and robustness during programmed tasks or motions, which is crucial for the safety of the mobility-impaired user. Due to various levels of the user's disability, the human-exoskeleton interaction forces and external perturbations are unpredictable and could vary substantially and…
▽ More
A significant challenge for the control of a robotic lower extremity rehabilitation exoskeleton is to ensure stability and robustness during programmed tasks or motions, which is crucial for the safety of the mobility-impaired user. Due to various levels of the user's disability, the human-exoskeleton interaction forces and external perturbations are unpredictable and could vary substantially and cause conventional motion controllers to behave unreliably or the robot to fall down. In this work, we propose a new, reinforcement learning-based, motion controller for a lower extremity rehabilitation exoskeleton, aiming to perform collaborative squatting exercises with efficiency, stability, and strong robustness. Unlike most existing rehabilitation exoskeletons, our exoskeleton has ankle actuation on both sagittal and front planes and is equipped with multiple foot force sensors to estimate center of pressure (CoP), an important indicator of system balance. This proposed motion controller takes advantage of the CoP information by incorporating it in the state input of the control policy network and adding it to the reward during the learning to maintain a well balanced system state during motions. In addition, we use dynamics randomization and adversary force perturbations including large human interaction forces during the training to further improve control robustness. To evaluate the effectiveness of the learning controller, we conduct numerical experiments with different settings to demonstrate its remarkable ability on controlling the exoskeleton to repetitively perform well balanced and robust squatting motions under strong perturbations and realistic human interaction forces.
△ Less
Submitted 7 May, 2021;
originally announced May 2021.
Evaluation of a 1-DOF Hand Exoskeleton for Neuromuscular Rehabilitation
Authors:
Xinalian Zhou,
Ashley Mont,
Sergei Adamovich
Abstract:
A low-cost 1-DOF hand exoskeleton for neuromuscular rehabilitation has been designed and assembled. It consists of a base equipped with a servo motor, an index finger part, and a thumb part, connected through three gears. The index part has a tri-axial load cell and an attached ring to measure the finger force. An admittance control scheme was designed to provide intuitive control and positive for…
▽ More
A low-cost 1-DOF hand exoskeleton for neuromuscular rehabilitation has been designed and assembled. It consists of a base equipped with a servo motor, an index finger part, and a thumb part, connected through three gears. The index part has a tri-axial load cell and an attached ring to measure the finger force. An admittance control scheme was designed to provide intuitive control and positive force amplification to assist the user's finger movement. To evaluate the effects of different control parameters on neuromuscular re-sponse of the fingers, we created an integrated exoskeleton-hand musculo-skeletal model to virtually simulate and optimize the control loop. The exo-skeleton is controlled by a proportional derivative controller that computes the motor torque to follow a desired joint angle of the index part, which is obtained from inverse kinematics of a virtual end-effector mass driven by the finger force. We conducted parametric simulations of the exoskeleton in action, driven by the user's closing and opening finger motion, with different proportional gains, end-effector masses, and other coefficients. We com-pared the interaction forces between the index finger and the ring in both passive and active modes. The best performing assistive controller can re-duce the force from around 1.45N (in passive mode) to only around 0.52N, more than 64% of reduction. As a result, the muscle activations of the flex-ors and extensors were reduced significantly. We also noted the admittance control scheme is versatile and can also provide resistance (e.g. for strength training) by simply increasing the virtual end-effect mass.
△ Less
Submitted 16 July, 2019;
originally announced July 2019.
