Jiabin Peng

Jiabin Peng

Auto-generated profile, based on publicly available information

Jiabin Peng

College of Electronic and Information Engineering, Nanjing University of Aeronautics and Astronautics, Nanjing, China

Research Focus & Metrics

Academic Career Length

4Years (2023-Present)

(Calculated once Scilit publication data is synced: current year − year of first publication + 1)

Representative Publications

View All Publications
16 July 2024 in IEEE Transactions on Industrial Informatics
Verified
Journal Article
Closed access
LZLijuan Zhang
Lijuan Zhang
WYWeiguo Yi
Weiguo Yi
HLHang Lin
Hang Lin
JPJiabin Peng
Jiabin Peng
PGPan Gao
Pan Gao
Autonomous navigation of multiple unmanned aerial vehicles (UAVs) serves as the foundation for their widespread applications in various fields. However, multi-UAV cooperative navigation is a challenging problem because it is difficult for UAVs to plan efficient paths while avoiding collisions with both neighboring UAVs and other obstacles. In this work, an efficient reinforcement learning (RL)-based cooperative navigation algorithm (RL-CN) is proposed to make optimal decisions in dense and dynamic environments. In RL-CN, the multi-UAV cooperative navigation is formulated as a Markov decision process and an enhanced RL method is proposed for continuous control of multiple UAVs. To solve the sparse reward problem, a group of reward functions are designed in reward shaping. Next, a staged-tuning with two subactor networks strategy is developed to accelerate the training process and improve the navigation performance. Then, an enhanced prioritized experience replay strategy is designed by considering both Temporal Difference (TD) error and connectivity trend. Thus, high-value transitions are more frequently selected to further enhance the training process. Finally, we conduct comprehensive simulation experiments and provide comparative results to substantiate the effectiveness and robustness of RL-CN algorithm.

9 February 2023 in Applied Sciences
Verified
Journal Article
Open access
PDF file
JPJiabin Peng
Jiabin Peng
LZLijuan Zhang
Lijuan Zhang
MFmingqiu fan
mingqiu fan
NZNan Zhao
Nan Zhao
LLLei Lei
Lei Lei
JXJiangcheng Xia
Jiangcheng Xia
As one of the key techniques used in the perception layer of the Industrial Internet of Things (IIoT), radio frequency identification (RFID) has been widely applied for object tracing, smart warehouse management, product line monitoring, etc. In most applications, conveyor belts are prevalently implemented to accelerate the sorting efficiency for goods management. However, in such a system, tags quickly go through the reader’s reading range resulting in constant changing of the tag set and limited participating time of moving tags. As a result, it poses more challenges to the tag identification problem in mobile systems than in traditional static applications. In this work, a novel admission-control-based dynamic query tree (ACDQT) protocol is proposed for fast-moving tag identification. In ACDQT, two main strategies are developed, i.e., multi-round admission control (MRAC) and dynamic query tree recognition (DQTR). In MRAC, the reading process of multiple rounds is analyzed, and the number of admitted tags in each round is optimized. Thus, the tag lost ratio is guaranteed, and the identification process can be effectively accelerated. In DQTR, colliding tags are grouped into multiple subsets with the help of consecutive colliding bits in tag responses. By constructing a dynamic query tree, the number of collision slots is greatly reduced, and the identification efficiency in a single round is improved significantly. With MRAC and DQTR, ACDQT can support higher tag flow rate in mobile systems than existing works. Both theoretical analyses and simulation results are presented to demonstrate the effectiveness of ACDQT.