Laser-Based Reactive Navigation for Multirotor Aerial Robots using Deep Reinforcement Learning.
Carlos SampedroHriday BavleAlejandro Rodriguez-RamosPaloma de la PuentePascual CampoyPublished in: IROS (2018)
Keyphrases
- reinforcement learning
- real robot
- robot control
- mobile robot
- robotic control
- obstacle avoidance
- service robots
- function approximation
- collision free
- humanoid robot
- robotic systems
- agent architecture
- aerial images
- multi robot
- unknown environments
- robot behavior
- autonomous robots
- robot navigation
- satellite images
- autonomous vehicles
- machine learning
- imitation learning
- temporal difference
- indoor environments
- state space
- robotic agents
- autonomous navigation
- navigation systems
- aerial imagery
- human robot
- action selection
- reactive planning
- formation control
- laser beam
- multi agent
- manipulation tasks
- function approximators
- lidar data
- optimal control
- model free
- neural network
- human robot interaction
- laser radar
- human users
- rough terrain
- dead reckoning
- learning algorithm
- fiber optic
- path planning
- multi robot systems
- vehicle tracking
- optimal policy
- multiple robots
- multi agent systems