Safe residual reinforcement learning for helicopter aerial refueling.
Damsara JayarathneSantiago PaternainSandipan MishraPublished in: AIM (2023)
Keyphrases
- reinforcement learning
- function approximation
- learning algorithm
- multi agent
- markov decision processes
- state space
- optimal policy
- transfer learning
- control problems
- robot control
- reinforcement learning algorithms
- model free
- aerial images
- mathematical model
- optimal control
- evolutionary algorithm
- relational reinforcement learning
- learning process
- temporal difference
- robotic control
- real time
- flight data
- policy search
- residual error
- fitted q iteration
- partially observable
- man made
- learning capabilities
- action selection
- urban areas
- dynamic programming
- information retrieval
- neural network