Autonomous navigation in complex environment is an important requirement for the design of a robot. Active SLAM (simultaneous localization and mapping) combining, which combine path planning with SLAM, is proposed to improve the ability of autonomous navigation in complex environment. In this paper, fully convolutional residual networks are used to recognize the obstacles to get depth image. The avoidance obstacle path is planned by Dueling DQN algorithm in the robot’s navigation; at the same time, the 2D map of the environment is built based on FastSLAM. The experiments show that the proposed algorithm can successfully identify and avoid moving and static obstacles with different quantities in the environment, and realize the autonomous navigation of the robot in a complex environment.
This is a preview of subscription content, log in to check access.
Buy single article
Instant unlimited access to the full article PDF.
Price includes VAT for USA
Subscribe to journal
Immediate online access to all issues from 2019. Subscription will auto renew annually.
This is the net price. Taxes to be calculated in checkout.
Chanier F, Checchin P, Blanc C, et al (2008) Map fusion based on a multi-map SLAM framework. In: 2008 IEEE international conference on multisensor fusion and integration for intelligent systems, 2008, MFI. IEEE, pp 533–538
Gouaillier D, Collette C, Kilner C (2010) Omni-directional closed-loop walk for NAO. In: 2010 10th IEEE-RAS international conference on humanoid robots (Humanoids). IEEE, pp 448–454
Chaves SM, Kim A, Eustice RM (2014) Opportunistic sampling-based planning for active visual SLAM. In: 2014 IEEE/RSJ international conference on intelligent robots and systems (IROS 2014). IEEE, pp 3073–3080
Prozorov AV, Priorov AL, Tyukin AL et al (2017) Algorithm for simultaneous localization and mapping based on video signal analysis. Meas Tech 59(10):1088–1093
Osswald S, Hornung A, Bennewitz M (2010) Learning reliable and efficient navigation with a humanoid. In: IEEE international conference on robotics and automation. IEEE, pp 2375–2380
Wei C, Xu J, Wang C, et al (2013) An approach to navigation for the humanoid robot Nao in domestic environments. In: Conference towards autonomous robotic systems. Springer, Berlin, pp 298–310
Fulgenzi C, Ippoliti G, Longhi S (2009) Experimental validation of FastSLAM algorithm integrated with a linear features based map. Mechatronics 19(5):609–616
Hornung A, Kai MW, Bennewitz M (2010) Humanoid robot localization in complex indoor environments. In: IEEE/RSJ international conference on intelligent robots and systems. IEEE, pp 1690–1695
Berns K, von Puttkamer E (2009) Simultaneous localization and mapping (SLAM). In: Autonomous land vehicles. Springer, Berlin, pp 146–172
Havangi R, Taghirad HD, Nekoui MA et al (2014) A square root unscented FastSLAM with improved proposal distribution and resampling. IEEE Trans Ind Electron 61(5):2334–2345
Kim DK, Chen T (2015) Deep neural network for real-time autonomous indoor navigation. ArXiv preprint arXiv:1511.04668
Giusti A, Guzzi J, Ciresan DC et al (2016) A machine learning approach to visual perception of forest trails for mobile robots. IEEE Robot Autom Lett 1(2):661–667
Tai L, Li S, Liu M (2016) A deep-network solution towards model-less obstacle avoidance. In: 2016 IEEE/RSJ international conference on intelligent robots and systems (IROS). IEEE, pp 2759–2764
Maček K, Petrović I, Perić N (2002) A reinforcement learning approach to obstacle avoidance of mobile robots. In: 7th international workshop on advanced motion control. pp 462–466
Zhou Y, Er MJ (2006) Self-learning in obstacle avoidance of a mobile robot via dynamic self-generated fuzzy Q-learning. In: 2006 and international conference on intelligent agents, web technologies and internet commerce, international conference on computational intelligence for modelling, control and automation. IEEE, pp 116–116
Laina I, Rupprecht C, Belagiannis V, et al (2016) Deeper depth prediction with fully convolutional residual networks. In: 2016 Fourth international conference on 3D vision (3DV). IEEE, pp 239–248
Wen S, Zheng W, Zhu J et al (2012) Elman fuzzy adaptive control for obstacle avoidance of mobile robots using hybrid force/position incorporation. IEEE Trans Syst Man Cybern Part C (Appl Rev) 42(4):603–608
Wen S, Chen X, Ma C et al (2015) The Q-learning obstacle avoidance algorithm based on EKF-SLAM for NAO autonomous walking under unknown environments. Robot Auton Syst 72:29–36
Xie L, Wang S, Markham A, et al (2017) Towards monocular vision based obstacle avoidance through deep reinforcement learning. ArXiv preprint arXiv:1706.09829
Wang Z, Schaul T, Hessel M, et al (2015) Dueling network architectures for deep reinforcement learning. ArXiv preprint arXiv:1511.06581
Gruslys A, Dabney W, Azar MG et al (2017) The reactor: a fast and sample-efficient actor-critic agent for reinforcement learning. ArXiv preprint arXiv:1704.04651v2
Chen J, Bai T, Huang X, et al (2017) Double-task deep Q-learning with multiple views. In: Proceedings of the IEEE international conference on computer vision. pp 1050–1058
This work was supported by the National Natural Science Foundation of China under Grant No. 61773333.
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
About this article
Cite this article
Wen, S., Zhao, Y., Yuan, X. et al. Path planning for active SLAM based on deep reinforcement learning under unknown environments. Intel Serv Robotics (2020). https://doi.org/10.1007/s11370-019-00310-w
- Path planning
- Deep reinforcement learning