Zun Liu, PhD, Assistant Professor/Distinguished Researcher, Peacock Plan Distinguished Talent, Doctoral Supervisor, Shenzhen University
Email: zunliu@szu.edu.cn
He received his Bachelor of Engineering degree in 2011 and his PhD in Engineering degree in 2019 from South China University of Technology. His research focuses on intelligent robot systems, unmanned aerial vehicles (UAVs), and reinforcement learning algorithms. He has led one National Natural Science Foundation of China (NSFC) general program, one National Natural Science Foundation of China (NSFC) youth program (completed), and one Shenzhen Municipal Natural Science Foundation project. He has also served as project manager and technical lead for one CGN key project, one National Key JKW project, and one key joint project. His research findings have been published in high-level international journals and conferences such as IEEE Transactions on Intelligent Transportation Systems and IEEE Transactions on Vehicular Technology.
He proposed the Drq deep reinforcement learning algorithm based on hierarchical structure and attention mechanism for autonomous navigation of UAVs, improving the intelligence level of unmanned autonomous systems and achieving significant progress. He developed a VIO-SLAM localization method based on point-line features and a laser-based SLAM localization method, effectively ensuring the localization accuracy of UAVs in GPS-free environments, thereby further expanding the application scenarios of UAVs. Given the current research status quo regarding the lack of universally applicable theories and methods for intelligent perception and decision-making in robotic systems, and the limited room for improvement in application scenarios and scope of application, future work will focus on complex indoor environments. This research proposes a knowledge-domain-based hierarchical reinforcement learning method for indoor rescue robot systems, aiming to construct an indoor intelligent rescue robot system with autonomous decision-making capabilities in specific scenarios. The research results have been applied to UAV defect detection in nuclear power plant containment structures. As the project manager and key technical leader, my main work included researching the design, developing algorithms, conducting field experiments, and developing a highly automated intelligent robot system for containment surface defect detection. This system enables rapid and automatic acquisition of containment surface images, eliminates the need for manual intervention, and possesses intelligent defect detection, key information extraction, and intelligent analysis capabilities. It has been successfully deployed and is operating at the CGN Daya Bay Nuclear Power Plant and Yangjiang Nuclear Power Plant.
Currently, my main research directions include intelligent robot systems, embodied intelligence key technologies, and intelligent UAV systems. Students are welcome to apply.
Papers: Author, Title, Journal Name, Volume (Issue) (Year), Page Numbers;
(1) Zun Liu; Yuanqiang Cao; Jianyong Chen; Jianqiang Li; A Hierarchical Reinforcement Learning Algorithm Based on Attention Mechanism for UAV Autonomous Navigation, IEEE Transactions on Intelligent Transportation Systems, 2023, 2022(1): 1-12 (Journal Article, CAS Zone 1, IF=8.5)
(2) Zun Liu; Jianqiang Li; Cheng Wang; Richard Yu; Jie Chen; Ying He; Changyin Sun; System Identification Based on Generalized Orthonormal Basis Function for Unmanned Helicopters: A Reinforcement Learning Approach, IEEE Transactions on Vehicular Technology, 2021, 70(2): 1135-1145 (Journal Article, CAS Zone 2, IF=6.8)
(3) Zun Liu; Xiaonan Hu; Jianqiang Li; Jie Chen; Wenlian Huang; Xiaoyu Zhao; Victor C.M.Leung; Graph relation network for person counting in construction site using UAV, Applied Soft Computing, 2021, 110: 107562 (Journal article, Chinese Academy of Sciences District 1, IF=8.7)
(4) Zun Liu; Ji, Honghai; Pei, Hailong*; Lewis, Frank L.; A new information-weighted recursive algorithm for time-varying systems: application to UAV system identification, International Journal of Systems Science, 2018, 49(11): 2477-2489. (Journal article, SCI included, IF=4.3)
(5) Jie Chen; Yifan Zhang; Jianqiang Li; Du Weiming; Zhuangzhuang Chen; Zun Liu; Huihui Wang; Victor C. M.Leung; Integrated Air-Ground Vehicles for UAV Emergency Landing Based on Graph Convolution Network, IEEE Internet of Things Journal, 2021, 9(12): 9106-9116. (Journal article, CAS Zone 1, IF=10.6)
（6）Zhuangzhuang Chen; Jin Zhang; Zhuonan Lai; Jie Chen; Zun Liu; Jianqiang Li; Geometry-Aware Guided Loss for Deep Crack Recognition, 2022 IEEE/CVF Conference on Computer Vision and Pattern Recognition (CVPR), New Orleans, 2022-06-19. (Conference paper, CFF-A category)
（7）Zhuangzhuang Chen; Jin Zhang; Zhuonan Lai; Guanming Zhu; Zun Liu; Jie Chen; Jianqiang Li. The Devil is in the Crack Orientation: A New Perspective for Crack Detection. Proceedings of the International Conference on Computer Vision. 2023. (Conference paper, CFF-A category)
（8）Chao Yao; Changkun Jiang; Zun Liu; Jie Chen; Jianqiang Li; Optimal Capacity Allocation and Caching Strategy for Multi-UAV Collaborative Edge Caching, IEEE_ARM_2021, 2021-07-03 (Conference Paper, EI Indexed)
（9）Cheng Wang; Jianqiang Li; Jie Chen; Heng Zhang; Li Wang; Zun Liu; Interpretable Respiratory Sound Analysis with Ensemble Knowledge Distillation, IEEE_ARM_2021, 2021-07-03 (Conference Paper, EI Indexed)
Research Projects:
（1）National Natural Science Foundation of China (NSFC) Project, 62373258, Research on Key Technologies for Decision Optimization of Indoor Rescue Robots, 2024-01-01 to 2027-12-31, RMB 500,000, Approved, Principal Investigator
（2）National Natural Science Foundation of China (NSFC) Youth Project, 62006157, Research on Multi-Agent Reinforcement Learning Based on Air-Ground Collaborative Robot System, 2021-01-01 to 2023-12-31, RMB 240,000, Completed, Principal Investigator
(3) Shenzhen Science and Technology Innovation Bureau Project, JCYJ20240813141628038, Research on Autonomous Perception and Decision-Making Algorithm for Robots in GPS-denied Environment, 2024-12 to 2027-12, RMB 300,000, Ongoing, Principal Investigator
(4) CGN Key Project, 3100121868, Intelligent Robot for Detecting Surface Defects in Nuclear Power Plant Containment, 2020.11 to 2023.08, RMB 2,579,000, Completed, Project Manager and Key Technical Leader
(5) National Key R&D Program Project, 2020-JCJQ-ZD-267-00, Key Technologies for Cognition and Collaboration in Autonomous Ground Unmanned Support Systems 2021-01 to 2023-12, RMB 18.7 million, Completed, Participant
(6) National Joint Fund Project, U2013201, Fundamental Theory and Key Technologies of Intelligent Aerial Robots for Nuclear Power Plant Safety, 2021-01-01 to 2024-12-31, RMB 2.83 million, Completed, Participant
