Compensation for Undefined Behaviors During Robot Task Execution by Switching Controllers Depending on Embedded Dynamics in RNN.
Kanata SuzukiHiroki MoriTetsuya OgataPublished in: IEEE Robotics Autom. Lett. (2021)
Keyphrases
- evolutionary robotics
- control software
- real robot
- mobile robot
- recurrent neural networks
- unknown environments
- nearest neighbor
- control system
- collective behavior
- vision system
- dynamical systems
- human robot interaction
- mutual adaptation
- robot control
- humanoid robot
- embedded systems
- robotic systems
- autonomous robots
- reinforcement learning
- multi robot
- legged robots
- physical constraints
- minority game
- trajectory tracking
- temporal behavior
- obstacle avoidance
- dynamic behavior
- simulated robot
- dynamic model
- modular robots
- neural network
- genetic algorithm
- wheeled mobile robot
- real time
- service robots
- multi agent
- visual servoing
- plan execution
- artificial neural networks
- state space
- control law
- position and orientation