检索结果-内蒙古大学图书馆

GCN-Transformer Autoencoder with Knowledge Distillation for Unsupervised Video Anomaly Detection

Journal of advanced Computational Intelligence and intelligent Informatics 2025年第3期29卷 659-667页

作者： Yan, Mingchao Xiong, Yonghua She, Jinhua School of Automation China University of Geosciences No.388 Lumo Road Hongshan District Hubei Wuhan430074 China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems No.388 Lumo Road Hongshan District Hubei Wuhan430074 China Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education No.388 Lumo Road Hongshan District Hubei Wuhan430074 China School of Engineering Tokyo University of Technology 1404-1 Katakura Hachioji Tokyo192-0982 Japan

Video anomaly detection is crucial in intelligent surveillance, yet the scarcity and diversity of abnormal events pose significant challenges for supervised methods. This paper presents an unsupervised framework that integrates graph attention networks (GATs) and Transformer architectures, combining masked autoencoders (MAEs) with self-distillation training. GATs are utilized to model spatial and inter-frame relationships, while Transformers capture long-range temporal dependencies, overcoming the limitations of traditional MAE and self-distillation approaches. The model employs a two-stage training process: first, a lightweight MAE combined with a GAT-Transformer fusion constructs a knowledge distillation module;second, the student autoencoder is optimized by integrating a graph convolutional autoencoder and a classification head to identify synthetic anomalies. We evaluate the proposed method on three representative datasets—ShanghaiTech Campus, UBnormal, and UCSD Ped2—and achieve promising results. © Fuji Technology Press Ltd.

关键词： Network coding

来源：评论

学校读者我要写书评

暂无评论

A repetitive-control method based on sliding-mode control for nonlinear systems with external disturbances 41

A repetitive-control method based on sliding-mode control fo...

引用

第41届中国控制会议

作者： Yibing Wang Min Wu Manli Zhang Luefeng Chen Jinhua She School of Automation China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education School of Engineering Tokyo University of Technology

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

This paper presents a two-dimensional(2 D) repetitive-control method based on sliding-mode control for a nonlinear system with external *** a sliding-mode controller has a powerful ability to handle nonlinearity and external disturbances during the sliding motion,the reference-tracking and disturbance-suppression problems of the original nonlinear system are decomposed into two subproblems:reference tracking for a linear system and disturbance suppression for a nonlinear modified-repetitive control system(MRCS) with external *** the first one,a repetitive controller constructed with two positive time-delay loops is employed to increase the flexibility of the system design and guarantee high accuracy reference tracking.A continuous-discrete 2 D model of the linear MRCS is established to describe continuous control and discrete learning by using the 2 D property of repetitive control.A sufficient stability condition,in which three parameters are used to adjust control and learning,is given to guarantee the stability of the linear *** design of the system is transformed into parameter *** optimization algorithm is used to yield the optimal *** the other,a sliding-mode controller is developed to eliminate the effect of both nonlinearity and external disturbances on the system.A linear-matrix-inequality(LMI)-based sufficient stability condition is derived to guarantee the stability of the sliding *** condition is used to design the sliding-mode controller,which has a simple structure and does not require any specific information of nonlinearity and external ***,a numerical example and a comparison demonstrate the validity and advantage of the presented method.

关键词： Repetitive control sliding-mode control two-dimensional model

来源：评论

学校读者我要写书评

暂无评论

Research on Urban Air Quality Prediction Method Based on the Fusion of Wide-area Spatiotemporal Distribution Data 26

Research on Urban Air Quality Prediction Method Based on the...

引用

26th International Conference on Industrial Technology, ICIT 2025

作者： Tao, Zixing Cao, Weihua Gan, Chao Wuhan China School of Future Technology School of Automation Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan China

ISBN: (纸本)9798331521950

Air quality data exhibit nonlinearity, sensitivity to environmental factors, and long-term dependencies. Numerous factors influence air quality, making accurate predictions based on a single-dimensional dataset impractical. This study proposes a method for urban air quality prediction that integrates wide-area spatiotemporal data, designed to address the unique characteristics of air quality datasets. First, relevant wide-area spatiotemporal data are selected, and their correlations with air quality are systematically analyzed. Second, a Long Short-Term Memory (LSTM) network-based Transformer model is utilized to capture the long-term dependencies in the air quality sequences. Finally, the model successfully generates hourly multi-step predictions for six air pollutants. The experimental results show that the proposed method outperforms the method that relies on air quality data alone for multi-step air quality prediction. © 2025 IEEE.

关键词： Air quality

来源：评论

学校读者我要写书评

暂无评论

Multi-model Coupled Marine Environment Model in Offshore Drilling Process 41

Multi-model Coupled Marine Environment Model in Offshore Dri...

引用

第41届中国控制会议

作者： Yujie Cao Sheng Du Chengda Lu Min Wu Jinhua She School of Automation China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education School of Engineering Tokyo University of Technology

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

Safe and efficient deep-sea drilling is inseparable from accurate and timely recognition of the ocean and geological *** refers to ocean dynamics,meteorology,hydrology,ecology,chemistry,ocean sound,light physical characteristics,ocean geology and geography.A complex ocean environment is one of the challenges facing ocean engineering at *** is difficult to establish an independent ocean or geological model for controller of the deep-sea drilling *** the face of complex and changeable ocean environment,modeling and analysis of wave environmental factors alone is not suitable for the control of deep-sea drilling *** paper analyzes the influence of various marine environmental factors on the offshore drilling platform and studies the internal relationship between each ***,the models of ocean breeze,wave and ocean current were designed to realize the simulation of ocean environment and parameterized description of ocean environment for deep-sea drilling *** has laid the foundation for the design of a deep-sea drilling control system.

关键词： Marine environment model Offshore drilling process Multi-model coupled

来源：评论

学校读者我要写书评

暂无评论

The Innovative Multi-Cavity Pneumatic Actuator: A Chemical Reaction-Driven Device for Generating Mechanical Crawling Motion

The Innovative Multi-Cavity Pneumatic Actuator: A Chemical R...

引用

Robotics, control and automation (ICRCA), International Conference on

作者： Junqi Mu Yiyang Wei Xiaofeng Zong School of Automation China University of Geosciences Wuhan China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Wuhan China

ISBN: (数字)9798331544577

ISBN: (纸本)9798331544584

The low manufacturing cost and fast response speed of pneumatic actuators have attracted widespread attention from researchers in the field of soft robotics. Biomimetic design, as one of the mainstream design concepts of pneumatic actuators, has provided researchers with a source of inspiration. Most existing actuators are based on biomimetic design. In this paper, a pneumatic soft actuator with a biomimetic worm-like design for crawling robots is proposed. This actuator is composed of multiple mesh-like air chambers, and its main parameters are optimized through finite element analysis, enabling it to perform various complex movements. Moreover, in most existing pneumatic actuators, the air source is supplied by air pumps, while the designed actuator in this paper is powered by chemical reactions. By controlling the types and ratios of raw materials, different chemical reactions occur, leading to periodic inhalation and exhalation, allowing the actuator to perform repetitive movements as intended. This driving method can effectively combine with the mechanical motion of the crawling robot, enhancing the autonomy and environmental adaptability of soft robots.

关键词： Pneumatic actuators Performance evaluation Gases Biomimetics Pneumatic systems Soft robotics Chemical reactions Raw materials Manufacturing Finite element analysis

来源：评论

学校读者我要写书评

暂无评论

A Survey on IMU-and-Vision-based Human Pose Estimation for Rehabilitation 41

A Survey on IMU-and-Vision-based Human Pose Estimation for R...

引用

第41届中国控制会议

作者： Yuan Niu Jinhua She Chi Xu School of Automation China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems School of Engineering Tokyo University of Technology Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

With the rapid development of human pose estimation,the human-pose-estimation method applied in rehabilitation therapy has attracted more and more attention in today's aging *** them,inertial measurement units(IMU) and computer vision are two significant methods for human pose *** mainly pay attention to the application of these two methods in the rehabilitation therapy *** this paper,we first summarize the research status of IMU and computer vision in the field of rehabilitation ***,we find that combining the two methods can produce better ***,the main problems of the current research and possible development directions are identified.

关键词： Human Pose Estimation IMU Computer Vision Rehabilitation Therapy

来源：评论

学校读者我要写书评

暂无评论

Hand-Motion Intention Recognition Based on One-Dimensional Convolutional Neural Network 41

Hand-Motion Intention Recognition Based on One-Dimensional C...

引用

第41届中国控制会议

作者： Hao Wu Feng Wang Juan Zhao Jinhua She School of Automation China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education School of Engineering Tokyo University of Technology

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

Surface electromyographic(sEMG) signals play an important role in human-computer interaction between patients and rehabilitation robots in rehabilitation *** paper shows the effectiveness of one-dimensional convolutional neural networks(1-D CNN) in estimating hand motion intentions from sEMG *** database 1 was employed to train the ***,the data from a single subject was used to train a model,and the accuracy was roughly maintained at 80%.Then,the whole dataset was used to train the network,and the recognition accuracy was improved to 84.16%.This shows that 1-D CNN effectively estimates hand movements and achieves high accuracy even for a small number of data.

关键词： Electromyographic hand motion intention Ninapro database one-dimensional convolutional neural network(1-D CNN)

来源：评论

学校读者我要写书评

暂无评论

Analysis and State Feedback control Bifurcation of Fractional-Order Cyclic Gene Regulatory Networks with Time Delay 41

Analysis and State Feedback Control Bifurcation of Fractiona...

引用

第41届中国控制会议

作者： Feng Liu Shujiang Sun Juan Zhao Feiyue Zhuo Hua O Wang School of Automation China University of Geosciences Engineering Research Center of Ministry of education for intelligent technology of earth exploration Hubei key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Department of Mechanical Engineering Boston University

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

In this paper,a fractional-order cyclic gene regulatory network model with time delay is ***,the total time delay of the system is selected as the bifurcation parameter,and the condition for the existence of Hopf bifurcation is derived by analyzing its characteristic *** is found that the time delay affects the stability of the system,and the order affects the position of the bifurcation *** the time delay is greater than the critical time delay,the system loses ***,a state feedback controller is designed for the unstable *** is proved that the control method has good control effect for the system instability caused by Hopf ***,the correctness of the theoretical derivation is verified by simulation.

关键词： Fractional-order cyclic gene regulatory network Hopf bifurcation feedback control time delay

来源：评论

学校读者我要写书评

暂无评论

A Multi-Time-Scale Prediction Method for Blast Furnace Gas Utilization Rate Based on Causality Analysis

A Multi-Time-Scale Prediction Method for Blast Furnace Gas U...

引用

2023 China automation Congress, CAC 2023

作者： Zhao, Guoyu An, Jianqi Guo, Yunpeng Yang, Chenglin He, Yong School of Future Technology China University of Geosciences China School of Automation China University of Geosciences China Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems China Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Wuhan China

ISBN: (纸本)9798350303759

Blast furnace operating parameters regulate the gas utilization rate (GUR), and different operating parameters affect the GUR on different time scales. However, the existing methods only analyze and model the prediction on a single-time scale, and the accuracy of the model needs to be improved. In this paper, a multi-time-scale prediction method of blast furnace gas utilization rate based on causality is proposed. Firstly, according to the production characteristics of the blast furnace, burden and blast supply are the operation means to regulate the blast furnace on different time scales. Then, this paper presents a blast furnace operation parameter extraction method fusing correlation coefficient and information entropy, and selects two most informative parameters to characterize the operation. On this basis, the causal analysis method is used to prove that the selected operation parameters affect the development trend of GUR. Finally, the long and short time scale prediction models of GUR are established, which are verified by the data of the actual blast furnace production process, improving the prediction accuracy of the GUR efficiently. © 2023 IEEE.

关键词： Blast furnaces

来源：评论

学校读者我要写书评

暂无评论

A Soft-Sensor Modeling Method Based on Gated Stacked Isomorphic Autoencoder 41

A Soft-Sensor Modeling Method Based on Gated Stacked Isomorp...

引用

第41届中国控制会议

作者： Jianjun Yang Xiaolu Xu Xiaohui Yin Wenxing Liu Yi Yuan Jian Peng School of Automation China University of Geosciences Hubei Key Laboratory of Advanced Control and Intelligent Automation for Complex Systems Engineering Research Center of Intelligent Technology for Geo-Exploration Ministry of Education Huaxin Cement Co. Ltd

ISBN: (数字)9789887581536

ISBN: (纸本)9781665482561

Soft sensing mainly studies the real-time prediction of some key performance indicators or quality variables in the actual production process,which has the role of guiding production in the actual production *** stacked autoencoder is a multi-layer autoencoder *** input variables will be encoded and decoded through each layer of autoencoders,and the obtained hidden features will be retained as the input of the next *** this way,high-level data features can be successfully learned from the input layer to the intermediate *** isomorphic autoencoder reconstruct an identical data input layer,and the reconstruction is performed by minimizing the error of the decoded data from the original input ***,for a soft measurement model,some data information may also reduce the accuracy or generalization of the *** order to overcome the above shortcomings,this paper proposes a gated stacked isomorphic autoencoder,which evaluates the contribution of each hidden layer feature through the gating unit,and then integrates the information of different hidden layers to complete the prediction and estimation of related main ***,the effectiveness and feasibility of the method are verified in practical industrial cases.

关键词： Deep learning soft sensor stacked autoencoder gating unit

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：