1
|
Luo S, Jiang M, Zhang S, Zhu J, Yu S, Dominguez Silva I, Wang T, Rouse E, Zhou B, Yuk H, Zhou X, Su H. Experiment-free exoskeleton assistance via learning in simulation. Nature 2024; 630:353-359. [PMID: 38867127 DOI: 10.1038/s41586-024-07382-4] [Citation(s) in RCA: 1] [Impact Index Per Article: 1.0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 09/04/2023] [Accepted: 04/03/2024] [Indexed: 06/14/2024]
Abstract
Exoskeletons have enormous potential to improve human locomotive performance1-3. However, their development and broad dissemination are limited by the requirement for lengthy human tests and handcrafted control laws2. Here we show an experiment-free method to learn a versatile control policy in simulation. Our learning-in-simulation framework leverages dynamics-aware musculoskeletal and exoskeleton models and data-driven reinforcement learning to bridge the gap between simulation and reality without human experiments. The learned controller is deployed on a custom hip exoskeleton that automatically generates assistance across different activities with reduced metabolic rates by 24.3%, 13.1% and 15.4% for walking, running and stair climbing, respectively. Our framework may offer a generalizable and scalable strategy for the rapid development and widespread adoption of a variety of assistive robots for both able-bodied and mobility-impaired individuals.
Collapse
Affiliation(s)
- Shuzhen Luo
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, USA
- Department of Mechanical Engineering, Embry-Riddle Aeronautical University, Daytona Beach, FL, USA
| | - Menghan Jiang
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, USA
| | - Sainan Zhang
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, USA
| | - Junxi Zhu
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, USA
| | - Shuangyue Yu
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, USA
| | - Israel Dominguez Silva
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, USA
| | - Tian Wang
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, USA
| | - Elliott Rouse
- Neurobionics Lab, Department of Robotics, Department of Mechanical Engineering, University of Michigan, Ann Arbor, MI, USA
| | - Bolei Zhou
- Department of Computer Science, University of California, Los Angeles, CA, USA
| | - Hyunwoo Yuk
- Department of Mechanical Engineering, Korea Advanced Institute of Science and Technology, Daejeon, South Korea
| | - Xianlian Zhou
- Department of Biomedical Engineering, New Jersey Institute of Technology, Newark, NJ, USA
| | - Hao Su
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, USA.
- Joint NCSU/UNC Department of Biomedical Engineering, North Carolina State University, Raleigh, University of North Carolina at Chapel Hill, Chapel Hill, NC, USA.
| |
Collapse
|
2
|
Xia Y, Wei W, Lin X, Li J. Optimization of Torque-Control Model for Quasi-Direct-Drive Knee Exoskeleton Robots Based on Regression Forecasting. SENSORS (BASEL, SWITZERLAND) 2024; 24:1505. [PMID: 38475041 DOI: 10.3390/s24051505] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Subscribe] [Scholar Register] [Received: 01/02/2024] [Revised: 02/20/2024] [Accepted: 02/22/2024] [Indexed: 03/14/2024]
Abstract
The choice of torque curve in lower-limb enhanced exoskeleton robots is a key problem in the control of lower-limb exoskeleton robots. As a human-machine coupled system, mapping from sensor data to joint torque is complex and non-linear, making it difficult to accurately model using mathematical tools. In this research study, the knee torque data of an exoskeleton robot climbing up stairs were obtained using an optical motion-capture system and three-dimensional force-measuring tables, and the inertial measurement unit (IMU) data of the lower limbs of the exoskeleton robot were simultaneously collected. Nonlinear approximations can be learned using machine learning methods. In this research study, a multivariate network model combining CNN and LSTM was used for nonlinear regression forecasting, and a knee joint torque-control model was obtained. Due to delays in mechanical transmission, communication, and the bottom controller, the actual torque curve will lag behind the theoretical curve. In order to compensate for these delays, different time shifts of the torque curve were carried out in the model-training stage to produce different control models. The above model was applied to a lightweight knee exoskeleton robot. The performance of the exoskeleton robot was evaluated using surface electromyography (sEMG) experiments, and the effects of different time-shifting parameters on the performance were compared. During testing, the sEMG activity of the rectus femoris (RF) decreased by 20.87%, while the sEMG activity of the vastus medialis (VM) increased by 17.45%. The experimental results verify the effectiveness of this control model in assisting knee joints in climbing up stairs.
Collapse
Affiliation(s)
- Yuxuan Xia
- School of Optoelectronic Science and Engineering, Soochow University, Suzhou 215031, China
| | - Wei Wei
- School of Electronic & Information Engineering, Suzhou University of Science and Technology, Suzhou 215009, China
| | - Xichuan Lin
- MebotX Intelligent Technology (Suzhou) Co., Ltd., Suzhou 215131, China
| | - Jiaqian Li
- School of Electronic & Information Engineering, Suzhou University of Science and Technology, Suzhou 215009, China
| |
Collapse
|
3
|
Yu S, Huang TH, Di Lallo A, Zhang S, Wang T, Fu Q, Su H. Bio-inspired design of a self-aligning, lightweight, and highly-compliant cable-driven knee exoskeleton. Front Hum Neurosci 2022; 16:1018160. [PMID: 36419645 PMCID: PMC9677347 DOI: 10.3389/fnhum.2022.1018160] [Citation(s) in RCA: 1] [Impact Index Per Article: 0.5] [Reference Citation Analysis] [Abstract] [Key Words] [Grants] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 08/12/2022] [Accepted: 10/19/2022] [Indexed: 04/07/2024] Open
Abstract
Powered knee exoskeletons have shown potential for mobility restoration and power augmentation. However, the benefits of exoskeletons are partially offset by some design challenges that still limit their positive effects on people. Among them, joint misalignment is a critical aspect mostly because the human knee joint movement is not a fixed-axis rotation. In addition, remarkable mass and stiffness are also limitations. Aiming to minimize joint misalignment, this paper proposes a bio-inspired knee exoskeleton with a joint design that mimics the human knee joint. Moreover, to accomplish a lightweight and high compliance design, a high stiffness cable-tension amplification mechanism is leveraged. Simulation results indicate our design can reduce 49.3 and 71.9% maximum total misalignment for walking and deep squatting activities, respectively. Experiments indicate that the exoskeleton has high compliance (0.4 and 0.1 Nm backdrive torque under unpowered and zero-torque modes, respectively), high control bandwidth (44 Hz), and high control accuracy (1.1 Nm root mean square tracking error, corresponding to 7.3% of the peak torque). This work demonstrates performance improvement compared with state-of-the-art exoskeletons.
Collapse
Affiliation(s)
- Shuangyue Yu
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, United States
| | - Tzu-Hao Huang
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, United States
| | - Antonio Di Lallo
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, United States
| | - Sainan Zhang
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, United States
- Department of Mechanical Engineering, City College of New York, New York, NY, United States
| | - Tian Wang
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, United States
| | - Qiushi Fu
- Mechanical and Aerospace Engineering, University of Central Florida, Orlando, FL, United States
- NeuroMechanical Systems Laboratory, Biionix (Bionic Materials, Implants & Interfaces) Cluster, University of Central Florida, Orlando, FL, United States
| | - Hao Su
- Lab of Biomechatronics and Intelligent Robotics, Department of Mechanical and Aerospace Engineering, North Carolina State University, Raleigh, NC, United States
- Joint North Carolina State University/The University of North Carolina Department of Biomedical Engineering, NC State University, Raleigh, NC, United States
- Joint North Carolina State University/The University of North Carolina Department of Biomedical Engineering, University of North Carolina at Chapel Hill, Chapel Hill, NC, United States
| |
Collapse
|