DRL Path Planning for UAV-Aided V2X Networks: Comparing Discrete to Continuous Action Spaces.
Leonardo SpampinatoAlessia TarozziChiara BurattiRiccardo MariniPublished in: ICASSP (2023)
Keyphrases
- path planning
- continuous action
- action space
- continuous state spaces
- continuous state
- state space
- mobile robot
- policy search
- dynamic environments
- path planning algorithm
- collision avoidance
- unmanned aerial vehicles
- real valued
- reinforcement learning
- obstacle avoidance
- markov decision processes
- motion planning
- path finding
- optimal path
- potential field
- multi robot
- single agent
- degrees of freedom
- aerial vehicles
- network structure
- autonomous vehicles
- stochastic processes
- dynamic and uncertain environments
- robot path planning
- search and rescue
- collision free
- configuration space
- skill learning
- partially observable markov decision processes
- belief state
- autonomous robots