Design and Implement an Enhanced Simulator for Autonomous Delivery Robot

Tian, Zhaofeng; Shi, Wei

doi:10.1109/metrocad56305.2022.00009

Cited by 6 publications

(2 citation statements)

References 10 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…The experiments are conducted on our ZebraT platform, a 1-meter long and 0.7-meter wide Ackermann-steering rectangular-shaped robot, and its dedicated ROS/Gazebo simulator [9]. While the training process is subject to the OpenAI gym framework and conducted in the Gazebo.…”

Section: Experiments and Resultsmentioning

confidence: 99%

“…We first incorporate our proposed methods into five mainstream DRL algorithms -DDPG, DQN, PPO, PPO-discrete, and SAC -and benchmark the corresponding performance in simulated narrow spaces. Specifically, we build upon our ZebraT platform [9] -a Lidar-based and rectangular-shaped robot using the Robot Operating System (ROS) -and simulate it in the Gazebo simulator environment on a narrow track comprised of straight corridors, 45-degree, 90-degree, and 180-degree corners. The results show that DDPG, DQN, and SAC accompanied by our proposed safety region and reward function all achieve good performance with DDPG being the best one.…”

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

Reinforcement Learning for Self-exploration in Narrow Spaces

Tian¹,

Liu²,

Zhou³

et al. 2022

Preprint

Self Cite

View full text Add to dashboard Cite

In narrow spaces, motion planning based on the traditional hierarchical autonomous system could cause collisions due to mapping, localization, and control noises. Additionally, it is disabled when mapless. To tackle these problems, we leverage deep reinforcement learning which is verified to be effective in self-decision-making, to self-explore in narrow spaces without a map while avoiding collisions. Specifically, based on our Ackermann-steering rectangular-shaped ZebraT robot and its Gazebo simulator, we propose the rectangular safety region to represent states and detect collisions for rectangular-shaped robots, and a carefully crafted reward function for reinforcement learning that does not require the destination information. Then we benchmark five reinforcement learning algorithms including DDPG, DQN, SAC, PPO, and PPO-discrete, in a simulated narrow track. After training, the well-performed DDPG and DQN models can be transferred to three brand new simulated tracks, and furthermore to three real-world tracks. https://sites.google.com/view/rl4exploration

show abstract

Section: Experiments and Resultsmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%