An advantage actor-critic algorithm for robotic motion planning in dense and dynamic scenarios.
Chengmin ZhouBingding HuangPasi FräntiPublished in: CoRR (2021)
Keyphrases
- motion planning
- objective function
- multi robot
- actor critic
- dynamic programming
- configuration space
- path planning
- robotic arm
- learning algorithm
- function approximation
- policy gradient
- dynamic environments
- linear programming
- mobile robot
- optimal solution
- reinforcement learning
- degrees of freedom
- humanoid robot
- negative matrix factorization
- gradient method