A Deep Reinforcement Learning Based Approach for AGVs Path Planning
chinese automation congress(2020)
Abstract
Autonomous path planning of automated guided vehicles (AGVs) is an important part of the intelligent logistics systems (ILS), which can greatly improve the abilities of intelligence and automation. Traditional AGVs navigation, electromagnetic navigation, and tape navigation, only navigate on metal belts or magnetic tapes, and the freedom of navigation is greatly reduced. This paper models the AGVs path planning problem as a reinforcement learning model to improve the capabilities of autonomous path planning and freedom of navigation. The path planning method based on the Dueling Double Deep Q Network with Prioritized Experience Reply (Dueling DDQN-PER) is used to learn the AGVs control in a simulation environment of an ILS. Using multi-modal sensory environment information, the AGV’s ability to approach the target location and avoid obstacles has been significantly improved. Experimental results show that the AGVs can autonomously plan the path in the environment with obstacles, and has a high success rate and obstacle avoidance ability.
MoreTranslated text
Key words
deep reinforcement learning, path planning, AGV
AI Read Science
Must-Reading Tree
Example
![](https://originalfileserver.aminer.cn/sys/aminer/pubs/mrt_preview.jpeg)
Generate MRT to find the research sequence of this paper
Chat Paper
Summary is being generated by the instructions you defined