Reinforcement learning based variable damping control of wearable robotic limbs for maintaining astronaut pose during extravehicular activity.
Sikai ZhaoTianjiao ZhengDongbao SuiJie ZhaoYanhe ZhuPublished in: Frontiers Neurorobotics (2023)
Keyphrases
- reinforcement learning
- control strategy
- control problems
- perception action
- robotic systems
- optimal control
- control system
- robot motion
- robot control
- robotic manipulator
- control loop
- d objects
- robotic arm
- pose estimation
- human body
- control method
- real robot
- open loop
- state space
- physical activity
- control policy
- adaptive control
- robotic control
- operating conditions
- machine learning
- function approximation
- human activities
- mobile robot
- human computer interaction
- hand eye
- autonomous robots
- real time
- learning algorithm
- computer vision
- dynamic programming
- lyapunov function
- vision system
- highly nonlinear
- optimal policy
- ambient intelligence
- mathematical model
- transfer learning
- human pose
- model free
- gesture recognition