Attention-Based Population-Invariant Deep Reinforcement Learning for Collision-Free Flocking with A Scalable Fixed-Wing UAV Swarm.
Chao YanKin Huat LowXiaojia XiangTianjiang HuLincheng ShenPublished in: IROS (2022)
Keyphrases
- collision free
- path planning
- reinforcement learning
- swarm intelligence
- dynamic environments
- collision avoidance
- mobile robot
- path planner
- motion planning
- potential field
- evolutionary process
- unmanned aerial vehicles
- particle swarm optimization
- state space
- free space
- multi robot
- real robot
- markov decision processes
- machine learning
- degrees of freedom
- spatio temporal
- viewpoint
- high dimensional
- control policy
- search algorithm
- three dimensional
- learning algorithm