Trading-Off Safety with Agility Using Deep Pose Error Estimation and Reinforcement Learning for Perception-Driven UAV Motion Planning.
Mehmetcan KaymazRecep AyzitOnur AkgünKamil Canberk AtikMustafa ErdemBaris YalcinGürkan CetinNazim Kemal UrePublished in: J. Intell. Robotic Syst. (2024)
Keyphrases
- motion planning
- error estimation
- path planning
- reinforcement learning
- degrees of freedom
- pose estimation
- mobile robot
- model selection
- trajectory planning
- robot arm
- error estimates
- humanoid robot
- multi robot
- robotic tasks
- robotic arm
- d objects
- autonomous mobile robot
- control law
- collision free
- inverse kinematics
- configuration space
- manipulation tasks
- generalization error
- obstacle avoidance
- collision avoidance
- state space
- optimal policy
- position and orientation
- dynamic environments
- machine learning
- learning algorithm
- data sets
- supervised learning
- similarity measure
- mechanical systems
- climbing robot