Abstract
In recent years, reinforcement learning, which enables robots to learn previously missing abilities, plays an increasingly important role in robotics, such as learning hard-to-code behaviors or optimizing problems without an accepted closed solution. The main problem of RL in robotics is that it is expensive and takes a long time to learn and operate. Another problem: advanced robot simulators like Gazebo are inefficient and time-consuming. In order to cope with these problems, a hybrid computing platform based on traditional robot simulation architecture and distributed architecture (hereinafter referred to as RDTP) is proposed in this paper, which helps to save cost, shorten time and speed up simulation and training. Additionally, the platform is optimized to a certain extent in terms of ease of use and compatibility.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Similar content being viewed by others
References
Brockman, G., et al.: OpenAI gym. arXiv preprint arXiv:1606.01540 (2016)
Fan, B., Pan, Q., Zhang, H.: A method for multi-agent coordination based on distributed reinforcement learning. Comput. Simul. 22(6), 115–117 (2005)
Zaharia, M., et al.: Resilient distributed datasets: a fault-tolerant abstraction for in-memory cluster computing. In: Proceedings of the 9th USENIX Conference on Networked Systems Design and Implementation, p. 2. USENIX Association, April 2012
Moritz, P., et al.: Ray: a distributed framework for emerging AI applications. In: 13th USENIX Symposium on Operating Systems Design and Implementation (OSDI 2018), pp. 561–577 (2018)
Tian, Y., Zhu, Y.: Better computer go player with neural network and long-term prediction. arXiv preprint arXiv:1511.06410 (2015)
Jouppi, N.P., et al.: In-datacenter performance analysis of a tensor processing unit. In: 2017 ACM/IEEE 44th Annual International Symposium on Computer Architecture (ISCA), pp. 1–12. IEEE, June 2017
LXDE Homepage. https://lxde.org/. Accessed 20 Sept 2018
VNC Homepage. https://www.realvnc.com/en/. Accessed 20 Sept 2018
Nginx Homepage. https://www.nginx.com/. Accessed 20 Sept 2018
Abadi, M., et al.: TensorFlow: a system for large-scale machine learning. In: OSDI, vol. 16, pp. 265–283, November 2016
Liang, E., et al.: Ray RLLib: a composable and scalable reinforcement learning library. arXiv preprint arXiv:1712.09381 (2017)
Docker Homepage. https://www.docker.com/. Accessed 20 Sept 2018
Wierstra, D., Schaul, T., Peters, J., Schmidhuber, J.: Natural evolution strategies. In: IEEE Congress on Evolutionary Computation, CEC 2008 (IEEE World Congress on Computational Intelligence), pp. 3381–3387. IEEE, June 2008
Mujoco Homepage. http://www.mujoco.org/. Accessed 20 Sept 2018
Torcs Homepage. http://torcs.sourceforge.net/index.php. Accessed 20 Sept 2018
Kempka, M., Wydmuch, M., Runc, G., Toczek, J., Jaśkowski, W.: VizDoom: a doom-based AI research platform for visual reinforcement learning. In: 2016 IEEE Conference on Computational Intelligence and Games (CIG), pp. 1–8. IEEE, September 2016
Schulman, J., Wolski, F., Dhariwal, P., Radford, A., Klimov, O.: Proximal policy optimization algorithms. arXiv preprint arXiv:1707.06347 (2017)
Acknowledgement
This work was supported by the National Natural Science Foundation of China under Grants (U1509210 and 61771793), the Chongqing Research Program of Basic Research and Frontier Technology (No. cstc2015jcyjA40026, No. cstc2016jcyjA0568), the Science and Technology Research Program of Chongqing Municipal Education Commission (No. KJ1711278, KJ1601129, KJ1501134), the Natural Science Foundation of Yongchuan Science and Technology Commission (No. Ycstc, 2016nc2002), and the Open Research Project of the State Key Laboratory of Industrial Control Technology, Zhejiang University, China (No. ICT170330, ICT1800413).
Author information
Authors and Affiliations
Corresponding author
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2019 Springer Nature Singapore Pte Ltd.
About this paper
Cite this paper
Jiang, YL., Huang, ZQ., Cao, JJ., Liu, Y., Ma, X., Huang, Y. (2019). Robot Simulation and Reinforcement Learning Training Platform Based on Distributed Architecture. In: Sun, F., Liu, H., Hu, D. (eds) Cognitive Systems and Signal Processing. ICCSIP 2018. Communications in Computer and Information Science, vol 1005. Springer, Singapore. https://doi.org/10.1007/978-981-13-7983-3_48
Download citation
DOI: https://doi.org/10.1007/978-981-13-7983-3_48
Published:
Publisher Name: Springer, Singapore
Print ISBN: 978-981-13-7982-6
Online ISBN: 978-981-13-7983-3
eBook Packages: Computer ScienceComputer Science (R0)