Safe Reinforcement Learning for LiDAR-based Navigation via Control Barrier Function.
Lixing SongLuke FerdererShaoen WuPublished in: ICMLA (2022)
Keyphrases
- reinforcement learning
- control policy
- control problems
- robot control
- optimal control
- function approximators
- autonomous vehicles
- point cloud
- piecewise linear
- high resolution
- optimal policy
- function approximation
- autonomous navigation
- action selection
- control method
- transition model
- underwater vehicles
- markov decision processes
- state space
- learning algorithm
- machine learning
- adaptive control
- robot navigation
- real time
- control strategy
- learning problems
- mobile robot
- outdoor environments
- control policies