Deep Reinforcement Learning to Acquire Navigation Skills for Wheel-Legged Robots in Complex Environments
Authors:
- Xi ChenXCXi Chen
- Ali GhadirzadehAGAli Ghadirzadeh
- John FolkessonJohn Folkesson
- Marten BjorkmanMBMarten Bjorkman
- Patric JensfeltPJPatric Jensfelt
- 1 October 2018
- conference paper
- Published by Institute of Electrical and Electronics Engineers (IEEE)
- p. 3110-3116
Abstract
Mobile robot navigation in complex and dynamic environments is a challenging but important problem. Reinforcement learning approaches fail to solve these tasks efficiently due to reward sparsities, temporal complexities and high-dimensionality of sensorimotor spaces which are inherent in such problems. We present a novel approach to train action policies to acquire navigation skills for wheel-legged robots using deep reinforcement learning. The policy maps height-map image observations to motor commands to navigate to a target position while avoiding obstacles. We propose to acquire the multifaceted navigation skill by learning and exploiting a number of manageable navigation behaviors. We also introduce a domain randomization technique to improve the versatility of the training samples. We demonstrate experimentally a significant improvement in terms of data-efficiency, success rate, robustness against irrelevant sensory data, and also the quality of the maneuver skills.Keywords
This publication has 9 references indexed in Scilit:
- Mastering the game of Go without human knowledgeNature, 2017
- Virtual-to-real deep reinforcement learning: Continuous control of mobile robots for mapless navigation2017
- Domain randomization for transferring deep neural networks from simulation to the real world2017
- Deep predictive policy training using reinforcement learning2017
- DeepLocoACM Transactions on Graphics, 2017
- Terrain-adaptive locomotion skills using deep reinforcement learningACM Transactions on Graphics, 2016
- Deep spatial autoencoders for visuomotor learning2016
- Human-level control through deep reinforcement learningNature, 2015
- V-REP: A versatile and scalable robot simulation framework2013