Hybrid autonomous controller for bipedal robot balance with deep reinforcement learning and pattern generators.
Christos KouppasMohamad SaadaQinggang MengMark KingDennis MajoePublished in: Robotics Auton. Syst. (2021)
Keyphrases
- autonomous learning
- reinforcement learning
- robotic systems
- humanoid robot
- autonomous robots
- control architecture
- walking speed
- real robot
- robot control
- motion control
- closed loop
- mobile robot
- optimal control
- service robots
- robotic manipulator
- semi autonomous
- position control
- autonomous navigation
- robot behavior
- search and rescue
- biped walking
- control system
- human robot interaction
- robot navigation
- real time
- motion planning
- internal representations
- biped robot
- control law
- autonomous vehicles
- force control
- control signals
- path planning
- vision system
- pid controller
- feedback control
- multi robot
- trajectory tracking
- perceptual aliasing
- central pattern generator
- inverted pendulum
- controller design
- temporal difference
- manipulation tasks
- robot manipulators
- obstacle avoidance
- function approximation
- control strategy
- state space
- end effector
- control strategies
- model free
- control scheme
- control algorithm
- neural network