检索结果-内蒙古大学图书馆

arXiv 2025年

作者： Zhu, Fei Liu, Yujing Liu, Wenzhuo Zhang, Zhaoxiang Centre for Artificial Intelligence and Robotics Hong Kong Institute of Science & Innovation CAS Hong Kong Key Laboratory of Systems and Control Academy of Mathematics and Systems Science CAS China State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of Automation CAS China School of Artificial Intelligence University of Chinese Academy of Sciences China

Continual learning, which aims to learn multiple tasks sequentially, has gained extensive attention. However, most existing work focuses on empirical studies, and the theoretical aspect remains under-explored. Recently, a few investigations have considered the theory of continual learning only for linear regressions, establishes the results based on the strict independent and identically distributed (i.i.d.) assumption and the persistent excitation on the feature data that may be difficult to verify or guarantee in practice. To overcome this fundamental limitation, in this paper, we provide a general and comprehensive theoretical analysis for continual learning of regression models. By utilizing the stochastic Lyapunov function and martingale estimation techniques, we establish the almost sure convergence results of continual learning under a general data condition for the first time. Additionally, without any excitation condition imposed on the data, the convergence rates for the forgetting and regret metrics are provided. Copyright © 2025, The Authors. All rights reserved.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

5G1M: Indoor Fingerprint Positioning Using a Single 5G Module

引用

IEEE Sensors Journal 2024年

作者： Wang, Changhao Jin, Xi Xia, Changqing Xu, Chi Sun, Yiming Duan, Yong Chinese Academy of Sciences Key Laboratory of Networked Control Systems Shenyang110016 China Chinese Academy of Sciences Shenyang Institute of Automation Shenyang110169 China Shenyang University of Technology Shenyang110870 China

Many critical applications necessitate indoor positioning, and the implementation of indoor positioning cannot be separated from various wireless signals. Among these, 5G-based indoor positioning is gaining increasing attention due to the wide coverage of 5G base stations and the stability of 5G signals. However, current positioning algorithms based on 5G signals require specialized measuring tools or information from telecommunication operators. Therefore, this paper proposes an indoor positioning algorithm, called 5G1M, that uses only a single 5G module and several commercial base stations already built outdoors. Besides these, 5G1M does not need any further assistance. We design 5G1M based on fingerprint positioning, which is less dependent on the environment and devices. First, 5G1M incorporates a Siamese network model with Ghost modules and squeeze-and-excitation blocks to reduce reliance on the scale of the fingerprint database and computational power while maintaining accuracy. Then, transfer learning is used to adapt to long-term environmental changes, and a trajectory-fitting method is employed to mitigate instantaneous signal fluctuations. The experimental result shows that in such a simplified positioning platform, 5G1M achieves a positioning error of 2.53 meters, improving accuracy by 29.7% compared to existing algorithms, and it also demonstrates faster convergence with fewer model parameters. © 2001-2012 IEEE.

关键词： Base stations

来源：评论

学校读者我要写书评

暂无评论

Secure State Estimation against Sparse Attacks on a Time-varying Set of Sensors

引用

IFAC-PapersOnLine 2023年第2期56卷 270-275页

作者： Zishuo Li Muhammad Umar B. Niazi Changxin Liu Yilin Mo Karl H. Johansson Department of Automation Tsinghua University Beijing 100084 China Division of Decision and Control Systems Digital Futures School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm SE-100 44 Sweden Laboratory for Information and Decision Systems Massachusetts Institute of Technology 77 Massachusetts Avenue Cambridge MA 02139 USA

This paper studies the problem of secure state estimation of a linear time-invariant (LTI) system with bounded noise in the presence of sparse attacks on an unknown, time-varying set of sensors. At each time, the attacker has the freedom to choose an arbitrary set of no more than p sensors and manipulate their measurements without restraint. To this end, we propose a secure state estimation scheme and guarantee a bounded estimation error irrespective of the attack signals subject to 2p-sparse observability and a mild, technical assumption that the system matrix has no degenerate eigenvalues. The proposed scheme comprises a design of decentralized observers for each sensor based on the local observable subspace decomposition. At each time step, the local estimates of sensors are fused by a median operator to obtain a secure estimation, which is then followed by a local detection-and-resetting process of the decentralized observers. The estimation error is shown to be upper-bounded by a constant which is determined only by the system parameters and noise magnitudes. Moreover, we design the detector threshold to ensure that the benign sensors never trigger the detector. The efficacy of the proposed algorithm is demonstrated by its application on a benchmark example of IEEE 14-bus system. We show that our proposed scheme can effectively tolerate sparse attacks on an unknown set of sensors, ensuring a bounded estimation error and effectively detecting and resetting the attacked sensors.

关键词： Estimation and filtering Design of fault tolerant/reliable systems Estimation and fault detection

来源：评论

学校读者我要写书评

暂无评论

Fast convergent on-board Li-ion battery state of charge estimation via back-forth moving horizon estimation

引用

IEEE Transactions on Power Electronics 2025年

作者： Liu, Zhihao Xiao, Yu Yuan, Yuan Xu, Xiaodong Luo, Biao Huang, Tingwen Ministry of Education School of Automation and Key Laboratory of Industrial Intelligence and Systems Central South University Changsha410083 China Shenzhen University of Advanced Technology Faculty of Computer Science and Control Engineering Shenzhen China

Accurate state of charge (SOC) estimation is crucial for the safe operation of lithium-ion batteries (LIBs), yet existing methods are limited by sensitivity to initial SOC guess or high computational complexity. To address these challenges, this study proposes an onboard back-forth moving horizon estimation (BFMHE) framework for SOC estimation, achieving high precision, robustness, and computational efficiency compatible with microcontroller implementation. An efficient linear equivalent circuit model (ECM) is established, where the open-circuit voltage-SOC relationship is approximated by piecewise linear functions. The proposed framework employs two MHE estimators that iteratively operate in opposite directions over a finite time horizon, which enables rapid convergence. Moreover, this study provides a rigorous proof of the stability of the BFMHE method for the battery system with current input. The proposed approach is validated experimentally under various operating conditions and with different initial SOC guesses. Experimental results show that the proposed onboard BFMHE method achieves a maximum estimation error of less than 1% and quickly converges to the true SOC value, demonstrating strong robustness to initial estimation errors. © 1986-2012 IEEE.

关键词： Lithium-ion batteries

来源：评论

学校读者我要写书评

暂无评论

Differentially Private Distributed Nonconvex Stochastic Optimization with Quantized Communication

arXiv

引用

arXiv 2024年

作者： Chen, Jialong Wang, Jimin Zhang, Ji-Feng Key Laboratory of Systems and Control Institute of Systems Science Academy of Mathematics and Systems Science Chinese Academy of Sciences Beijing100190 China School of Mathematical Sciences University of Chinese Academy of Sciences Beijing100049 China School of Automation and Electrical Engineering University of Science and Technology Beijing Beijing100083 China Key Laboratory of Knowledge Automation for Industrial Processes Ministry of Education Beijing100083 China School of Automation and Electrical Engineering Zhongyuan University of Technology Zheng Zhou450007 China Key Laboratory of Systems and Control Institute of Systems Science Academy of Mathematics and Systems Science Chinese Academy of Sciences Beijing100190 China

This paper proposes a new distributed nonconvex stochastic optimization algorithm that can achieve privacy protection, communication efficiency and convergence simultaneously. Specifically, each node adds general privacy noises to its local state to avoid information leakage, and then quantizes its noise-perturbed state before transmitting to improve communication efficiency. By using a subsampling method controlled through the sample-size parameter, the proposed algorithm reduces cumulative differential privacy parameters ϵ, δ, and thus enhances the differential privacy level, which is significantly different from the existing works. By using a two-time-scale step-sizes method, the mean square convergence for nonconvex cost functions is given. Furthermore, when the global cost function satisfies the Polyak-Lojasiewicz condition, the convergence rate and the oracle complexity of the proposed algorithm are given. In addition, the proposed algorithm achieves both the mean square convergence and finite cumulative differential privacy parameters ϵ, δ over infinite iterations as the sample-size goes to infinity. A numerical example of the distributed training on the "MNIST" dataset is given to show the effectiveness of the algorithm. Copyright © 2024, The Authors. All rights reserved.

关键词： Differential privacy

来源：评论

学校读者我要写书评

暂无评论

Reliable deep learning in anomalous diffusion against out-of-distribution dynamics

NATURE COMPUTATIONAL SCIENCE

引用

NATURE COMPUTATIONAL SCIENCE 2024年第11期4卷 877-877页

作者： Feng, Xiaochen Sha, Hao Zhang, Yongbing Su, Yaoquan Liu, Shuai Jiang, Yuan Hou, Shangguo Han, Sanyang Ji, Xiangyang School of Computer Science and Technology Harbin Institute of Technology (Shenzhen) Shenzhen China Shenzhen International Graduate School Tsinghua University Shenzhen China Institute of Systems and Physical Biology Shenzhen Bay Laboratory Shenzhen China Department of Automation Tsinghua University Beijing China

Anomalous diffusion plays a crucial rule in understanding molecular-level dynamics by offering valuable insights into molecular interactions, mobility states and the physical properties of systems across both biological and materials sciences. Deep-learning techniques have recently outperformed conventional statistical methods in anomalous diffusion recognition. However, deep-learning networks are typically trained by data with limited distribution, which inevitably fail to recognize unknown diffusion models and misinterpret dynamics when confronted with out-of-distribution (OOD) scenarios. In this work, we present a general framework for evaluating deep-learning-based OOD dynamics-detection methods. We further develop a baseline approach that achieves robust OOD dynamics detection as well as accurate recognition of in-distribution anomalous diffusion. We demonstrate that this method enables a reliable characterization of complex behaviors across a wide range of experimentally diverse systems, including nicotinic acetylcholine receptors in membranes, fluorescent beads in dextran solutions and silver nanoparticles undergoing active endocytosis.

关键词： Silver nanoparticles

来源：评论

学校读者我要写书评

暂无评论

Indoor Fingerprint Positioning Method Based on Real 5G Signals 23

Indoor Fingerprint Positioning Method Based on Real 5G Signa...

引用

7th International Conference on Machine Learning and Soft Computing, ICMLSC 2023

作者： Wang, Changhao Xi, Jin Xia, Changqing Xu, Chi Duan, Yong Shenyang University of Technology Shenyang110870 China Key Laboratory of Networked Control Systems Chinese Academy of Sciences Shenyang110016 China Shenyang Institute of Automation Chinese Academy of Sciences Shenyang110016 China Institutes for Robotics and Intelligent Manufacturing Chinese Academy of Sciences Shenyang110169 China

ISBN: (纸本)9781450398633

Indoor positioning services are being used more and more widely. However, existing indoor positioning techniques cannot simultaneously take into account low cost, ease of use, high precision, and seamless switching between indoor and outdoor positioning. With the maturity of 5G techniques, 5G-based indoor positioning is gradually being paid attention to. 5G-based indoor positioning does not require additional equipment, and supports flexible indoor and outdoor switching under the same system. However, the 5G-related information used in existing research on 5G indoor positioning is not open to users. Therefore, in this paper, we propose an indoor fingerprint positioning method based on measured 5G signals. This method first collects 5G signals in the positioning area, and processes them to form a fingerprint database. Then, a machine learning algorithm is used to match the signal to be located with the fingerprint database to obtain the positioning result. Finally, we conduct experiments in real field, and the experimental result demonstrates that the positioning accuracy of our proposed method can reach 96%. © 2023 ACM.

关键词： 5G mobile communication systems

来源：评论

学校读者我要写书评

暂无评论

Iterative assist-as-needed control with interaction factor for rehabilitation robots

引用

Science China(Technological Sciences) 2021年第4期64卷 836-846页

作者： CAO Ran CHENG Long YANG ChenGuang DONG Zhe School of Artificial Intelligence University'of Chinese Academy of SciencesBeijing 100049China State Key Laboratory for Management and Control of Complex Systems Institute of Automation.Chinese Academy of SciencesBeijing 100190China School of Automation Science and Engineering South China University of TechnologyGuangzhou 510641China School of Electrical and Control Engineering North China Universityof TechnologyBeijing 100144China

Rehabilitation robots for stroke patients have drawn considerable attention because they can reduce the economic and labor costs brought by traditional rehabilitation. control methods for rehabilitation robots have been developed to stimulate the active motion of patients and to improve the effectiveness of rehabilitation care. However, current control methods can only roughly adjust the system’s stiffness and may fail in achieving satisfactory performance. To this end, this paper introduces a novel cost function consisting of the tracking error term and the stiffness term. The cost function contains an interaction factor that represents the patient’s motion intention to balance the weight of these two terms. When the patients try to actively do training tasks, the weight of stiffness term increases, which leads to the larger allowable tracking error and lower stiffness eventually. An iterative updating law of the stiffness matrix is given to reduce the proposed cost function. Theoretical analysis based on the Lyapunov theory is given to ensure the feasibility of the proposed algorithm. Furthermore, a force estimation is used to improve interaction control performance. Finally, simulation experiments are provided to show the effectiveness of the proposed algorithm.

关键词： human-robot interaction impedance stiffness control rehabilitation

来源：评论

学校读者我要写书评

暂无评论

A Novel Agricultural Data Sharing Mode Based on Rice Disease Identification

引用

Plant Diseases and Pests 2024年第2期15卷 9-16页

作者： Mengmeng ZHANG Xiujuan WANG Mengzhen KANG Jing HUA Haoyu WANG Feiyue WANG State Key Laboratory of Multimodal Artificial Intelligence Systems Institute of AutomationChinese Academy of SciencesBeijing100190China School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100049China Beijing Engineering Research Center of Intelligent Systems and Technology Institute of AutomationChinese Academy of SciencesBeijing 100190China Faculty of Innovation Engineering Macao University of Science and TechnologyMacao 999078China State Key Laboratory for Management and Control of Complex System Institute of AutomationChinese Academy of SciencesBeijing 100190China

In this paper,a variety of classical convolutional neural networks are trained on two different datasets using transfer learning *** demonstrated that the training dataset has a significant impact on the training results,in addition to the optimization achieved through the model ***,the lack of open-source agricultural data,combined with the absence of a comprehensive open-source data sharing platform,remains a substantial *** issue is closely related to the difficulty and high cost of obtaining high-quality agricultural data,the low level of education of most employees,underdeveloped distributed training systems and unsecured data *** address these challenges,this paper proposes a novel idea of constructing an agricultural data sharing platform based on a federated learning(FL)framework,aiming to overcome the deficiency of high-quality data in agricultural field training.

关键词： Rice disease and pest identification Convolutional neural networks Distributed training Federated learning(FL) Open-source data sharing platform

来源：评论

学校读者我要写书评

暂无评论

MALT: Multi-scale Action Learning Transformer for Online Action Detection

MALT: Multi-scale Action Learning Transformer for Online Act...

引用

International Joint Conference on Neural Networks (IJCNN)

作者： Zhipeng Yang Ruoyu Wang Yang Tan Liping Xie The Key Laboratory of Measurement and Control of Complex Systems of Engineering Ministry of Education School of Automation Southeast University Nanjing China Southeast University Shenzhen Research Institute Shenzhen China

ISBN: (数字)9798350359312

ISBN: (纸本)9798350359329

Online action detection (OAD) aims to identify ongoing actions from streaming video in real-time, without access to future frames. Since these actions manifest at varying scales of granularity, ranging from coarse to fine, projecting an entire set of action frames to a single latent encoding may result in a lack of local information, necessitating the acquisition of action features across multiple scales. In this paper, we propose a multi-scale action learning transformer (MALT), which includes a novel recurrent decoder (used for feature fusion) that includes fewer parameters and can be trained more efficiently. A hierarchical encoder with multiple encoding branches is further proposed to capture multi-scale action features. The output from the preceding branch is then incrementally input to the subsequent branch as part of a cross-attention calculation. In this way, output features transition from coarse to fine as the branches deepen. We also introduce an explicit frame scoring mechanism employing sparse attention, which filters irrelevant frames more efficiently, without requiring an additional network. The proposed method achieved state-of-the-art performance on two benchmark datasets (THUMOS’14 and TVSeries), outperforming all existing models used for comparison, with an mAP of 0.2% for THUMOS’14 and an mcAP of 0.1% for TVseries.

关键词： Filters Neural networks Streaming media Transformers Feature extraction Encoding Real-time systems

来源：评论

学校读者我要写书评

暂无评论

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案：

请选择收藏分类：

通借通还

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：