Musculoskeletal modeling and humanoid control of robots based on human gait data

• Main Authors: Jun Yu, Shuaishuai Zhang, Aihui Wang, Wei Li, Lulu Song
• Format: Article
• Language: English
• Published: PeerJ Inc. 2021-08-01
• Series: PeerJ Computer Science
• Subjects: Human-gait; Musculoskeletal model; OpenSim; Humanoid control; Lower limb rehabilitation exoskeleton; Motion capture
• Online Access: https://peerj.com/articles/cs-657.pdf

The emergence of exoskeleton rehabilitation training has brought good news to patients with limb dysfunction. Rehabilitation robots are used to assist patients with limb rehabilitation training and play an essential role in promoting the patient’s sports function with limb disease restoring to daily life. In order to improve the rehabilitation treatment, various studies based on human dynamics and motion mechanisms are still being conducted to create more effective rehabilitation training. In this paper, considering the human biological musculoskeletal dynamics model, a humanoid control of robots based on human gait data collected from normal human gait movements with OpenSim is investigated. First, the establishment of the musculoskeletal model in OpenSim, inverse kinematics, and inverse dynamics are introduced. Second, accurate human-like motion analysis on the three-dimensional motion data obtained in these processes is discussed. Finally, a classic PD control method combined with the characteristics of the human motion mechanism is proposed. The method takes the angle values calculated by the inverse kinematics of the musculoskeletal model as a benchmark, then uses MATLAB to verify the simulation of the lower extremity exoskeleton robot. The simulation results show that the flexibility and followability of the method improves the safety and effectiveness of the lower limb rehabilitation exoskeleton robot for rehabilitation training. The value of this paper is also to provide theoretical and data support for the anthropomorphic control of the rehabilitation exoskeleton robot in the future.