Policy Rollout Action Selection in Continuous Domains for Sensor Path Planning.
Folker HoffmannAlexander CharlishMatthew RitchieHugh D. GriffithsPublished in: IEEE Trans. Aerosp. Electron. Syst. (2021)
Keyphrases
- action selection
- path planning
- continuous domains
- reinforcement learning
- basal ganglia
- mobile robot
- robot soccer
- dynamic environments
- collision avoidance
- path planning algorithm
- evolutionary computation
- multi robot
- obstacle avoidance
- decision making
- temporal difference
- motion planning
- optimal path
- evolution strategy
- ground vehicles
- dynamic and uncertain environments
- genetic algorithm
- multiple robots
- action space
- unmanned aerial vehicles
- action selection mechanism
- higher dimensional
- optimal policy
- policy iteration
- markov decision process
- aerial vehicles
- vision system
- search and rescue
- state space
- real robot