作者：Weimin Liu^1,2;Shiqi Gao³;Feng Xu²;Yandong Zhao²;Yuanqing Xia¹;CA1;Jinkun Liu³; （¹School of Automation, Beijing Institute of Technology, China;²North Automatic Control Technology Institute, China;³School of Automation Science and Electrical Engineering, Beihang University, China）

出处：Communications in Nonlinear Science and Numerical Simulation 2023

摘要：Modeling and consensus control of flexible wings with bending deformation and torsion deformation are studied in this paper. Due to the physical prope ...

作者：Amit Krishan Kumar^1,2;CA1;Nguyễn Ngọc Mai³;Ke Tian⁴;Yuanqing Xia⁴; （¹Institute of Research and Development, Duy Tan University, 550000, Da Nang, Vietnam;²Faculty of Electrical-Electronic Engineering, Duy Tan University, 550000, Da Nang, Vietnam;³Faculty of Chinese Language, School of Foreign Language, Duy Tan University, Da Nang, Vietnam;⁴State Key Laboratory of Intelligent Control and Decision of Complex Systems, School of Automation, Beijing Institute of Technology, 100081, Beijing, China）

出处：SN Computer Science 2023

关键词：Isochronous;Nearest neighbour;Classification;Distance metric;Clustering

摘要：Machine learning, classification, and clustering techniques use the distance functions to evaluate the proximity between data entries and deduce the b ...

作者：闫莉萍¹;，李程伟¹;，徐柏凯¹;，夏元清¹;，肖波²; （¹北京理工大学自动化学院；²北京邮电大学人工智能学院）

出处：系统工程与电子技术 2023

关键词：同时定位与建图；点特征；线特征；结构特征；惯性测量单元

摘要：针对传统的仅依赖点特征的视觉里程计鲁棒性较差的问题，提出了一种结合点特征、线特征以及结构特征的视觉-惯性里程计。结构特征是一个满足曼哈顿世界假设的正交基，可用于描述结构物体包含的平行与正交信息。结构特征的引入为视觉惯性里程计提供了一个全局的约束，可有效减少旋转误差的累计。与此同时，为了保证引入结构特 ...

作者：董万城¹;，卜浩然¹;，江镇¹;，余思遥¹;，张立新¹;，夏元清²; （¹石河子大学机械电气工程学院；²北京理工大学自动化学院复杂系统智能控制与决策重点实验室）

出处：石河子大学学报(自然科学版) 2023

关键词：排肥器；槽型输送带式；锥齿轮；离散元

摘要：为实现排肥器的稳定排肥和排肥流量可调控，以及排肥器的创新设计与理论研究提供参考，设计一种改善排肥流量均匀性的槽型输送带式排肥器。对槽型输送带式排肥器的关键部件进行了优化设计，确定了排肥器小锥齿轮模数为3,齿数14、齿宽11 mm,并利用EDEM离散元软件对排肥过程进行仿真分析，以槽型同步带开度、锥齿 ...

作者：Sheng Yu¹;Di-Hua Zhai^1,2;Yuyin Guan³;Yuanqing Xia¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China ²Yangtze Delta Region Academy Beijing Institute of Technology, Jiaxing, China ³Beijing Building Materials Academy of Science Research, Beijing, China）

出处：IEEE Transactions on Neural Networks and Learning Systems 2023

关键词：Pose estimation;Feature extraction;Point cloud compression;Deformation;Shape;Transformers;Solid modeling

摘要：Category-level 6-D object pose estimation plays a crucial role in achieving reliable robotic grasp detection. However, the disparity between synthetic ...

作者：Yongkang Wang¹;Di-Hua Zhai^1,2;Yuanqing Xia¹;Danyang Liu¹; （¹School of Automation, Beijing Institute of Technology, Beijing, China ²Yangtze Delta Region Academy Beijing Institute of Technology, Jiaxing, China）

出处：IEEE Transactions on Industrial Informatics 2023

关键词：Training;Data models;Clustering algorithms;Internet of Things;Informatics;Electronic learning;Behavioral sciences

摘要：Robustness and attacks have become prominent concerns in federated learning (FL)-based Internet of Things (IoT). Our focus primarily lies on robustnes ...

发明人：邹伟东，刘天昊，夏元清

申请日期：2023.07.19

摘要：本发明公开了一种基于线性二次型调节器的极限学习机的图像识别方法，利用卷积运算随机提取图像特征与利用线性二次型调节器求取隐含层节点的输出权值矩阵，实现对极限学习机模型的训练，能够在降低计算负担过重和提高计算效率的情况下提高算法的泛化性能，在一定程度上能够满足图像识别的需要，同时为更准确地进行图像识别提 ...

发明人：邹伟东，卢辉煌，夏元清

申请日期：2023.08.16

摘要：本发明公开了一种基于变步长ADMM算法极限学习机的图像分类方法，提出了变步长交替方向乘子法(Variable Step Alternating Direction Method of Multipliers，Var‑ADMM)，通过动态调整求解过程的迭代步长，在弥补ADMM算法收敛缓慢的缺陷的同时有 ...

发明人：吴楚格，刘嘉宁，张若辰，夏元清，孙中奇，李怡然，詹玉峰

申请日期：2023.07.27

摘要：本发明公开了一种基于模型迁移学习的电子产品零件备货量预测方法充分利用历史数据，结合分位数回归的方法建立神经网络预测模型，通过对预测模型的不同分位点回归得到多个解与真实的历史数据比较，选择最优估计的分位点，再通过以均方根作为损失函数的神经网络一般建立方法进行验证，此外，对于无法收集到足够的需求样本数据 ...

作者：Jiaping Qiang¹;, Li Li ^{1 na1}; & Yuanqing Xia^{2 na1}; （¹Engineering Research Center of the Ministry of Education for Intelligent Control System and Intelligent Equipment, School of Electrical Engineering, Yanshan University, Qinhuangdao, 066004, Hebei Province, China;²School of Automation, Beijing Institute of Technology, Beijing, 100081, China）

出处：Nonlinear Dynamics

关键词：Prescribed-time control;Distributed control;Consensus control;High-order lower triangular nonlinear MASs

摘要：A novel distributed prescribed-time leader–follower tracking consensus control strategy is proposed for high-order nonlinear multi-agent systems with ...