Q-learning with experience replay in a dynamic environment.
Mathijs PietersMarco A. WieringPublished in: SSCI (2016)
Keyphrases
- dynamic environments
- potential field
- single agent
- reinforcement learning algorithms
- autonomous agents
- reinforcement learning
- mobile robot
- path planning
- learning algorithm
- state space
- cooperative
- collision avoidance
- function approximation
- changing environment
- stochastic approximation
- learning rate
- model free
- adaptive control
- agent systems
- real environment
- optimal policy
- environmental conditions