I will talk about my research on neuromechanical simulation models that can produce human-like diverse locomotion and their usage in studying human physiology, testing assistive devices, and controlling robotic systems. In my Ph.D. research, I proposed a reflex-based control model that could generate diverse human-like motions with a musculoskeletal model in physics simulations. Since then, we have adapted the model to explain elderly gait, predict the performance of gait assistive ankle exoskeletons, and control a bipedal robot. In addition to presenting these previous works, I will discuss how we plan to enhance the versatility of neuromechanical simulations using deep reinforcement learning and validate the reliability of their predictions, especially for gait assistive exoskeletons. 

• Neuromechanical simulation
• Locomotion
• Motor control
• Exoskeletons
• Deep reinforcement learning

목록

  전체댓글 0

[로그인]

댓글 입력란

0/500자