检索结果-内蒙古大学图书馆

Investigating the Impact of Hyper Parameters on Intrusion Detection system Using Deep Learning Based Data Augmentation

引用

International Journal of Advanced Computer Science and Applications 2025年第4期16卷 523-532页

作者： Iftikhar, Umar Ali, Syed Abbas Department of Computer & Information System Engineering NED University of Engineering & Technology Karachi Pakistan

The effects of changing learning rates, data augmentation percentage and numbers of epochs on the performance of Wasserstein Generative Adversarial Networks with Gradient Penalties (WGAN-GP) are evaluated in this study. The purpose of this research is to find out how they affect the data augmentation to enhance stability during training. In this research, the degree of system performance is measured using the Classification Model Utility approach. For this reason, this study aims to determine the interaction between learning rate, augmentation percentage and epoch value when using WGAN-GP to generate synthetic data for the recognition of the system performance. The results will provide the indications on how some of the hyper parameters can be adjusted up or down for having positive or negative consequences on the generation process for further research and use of WGAN-GP. It also provides insights into how the generative model is trained, and how that affects stability and quality of the result in various settings such as image synthesis or other generative tasks. © (2025), (Science and Information Organization). All rights reserved.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Kernel-Based State-Space Kriging for Predictive Control

引用

IEEE/CAA Journal of Automatica Sinica 2023年第5期10卷 1263-1275页

作者： A.Daniel Carnerero Daniel R.Ramirez Daniel Limon Teodoro Alamo the Department of System Engineering and Automation University of SevilleSevilla 41020Spain the School of Engineering Tokyo Institute of TechnologyTokyo 152-8552Japan Department of System Engineering and Automation University of SevilleSevilla 41020Spain

In this paper, we extend the state-space kriging(SSK) modeling technique presented in a previous work by the authors in order to consider non-autonomous systems. SSK is a data-driven method that computes predictions as linear combinations of past outputs. To model the nonlinear dynamics of the system, we propose the kernel-based state-space kriging(K-SSK), a new version of the SSK where kernel functions are used instead of resorting to considerations about the locality of the data. Also, a Kalman filter can be used to improve the predictions at each time step in the case of noisy measurements. A constrained tracking nonlinear model predictive control(NMPC) scheme using the black-box input-output model obtained by means of the K-SSK prediction method is proposed. Finally, a simulation example and a real experiment are provided in order to assess the performance of the proposed controller.

关键词： Data-driven methods model identification Kernel methods predictive control

来源：评论

学校读者我要写书评

暂无评论

Design of Lightweight Fully-Connected Network in Hardware Using Learning-Based Low-Rank Approximation and Quantization Techniques

引用

Transactions of the Korean Institute of Electrical Engineers 2025年第1期74卷 149-163页

作者： Seo, Jeong-Yun Lee, Jong-Youn Park, Sung-Jun Lee, Harim School of Electronic Engineering Kumoh National Institute of Technology Korea Republic of Department of Semiconductor System Engineering Kumoh National Institute of Technology Korea Republic of

In this paper, we address the design of an AI hardware accelerator optimized for a lightweight fully-connected network. Techniques such as quantization, knowledge distillation, pruning, and low-rank approximation are utilized to reduce the number of weights, maintaining inference performance while minimizing memory requirements. We introduce a learning-based low-rank approximation that outperforms the original low rank approximation. In addition, the interrelationship between various compression techniques such as quantization, knowledge distillation, pruning, and low-rank approximation is analyzed to enhance the understanding of deep learning model compression. In order to use the decomposed weight matrices in hardware, we design a compressed fully-connected layer, utilized to construct a lightweight fully-connected network. The proposed hardware design is developed by using Verilog HDL and verified through RTL simulation. Copyright © The Korean Institute of Electrical Engineers.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Hardware Implementation of Convolution Layer for Quantized Deep Neural Networks in Image Classification

引用

Transactions of the Korean Institute of Electrical Engineers 2025年第4期74卷 644-654页

作者： Lee, Jong-Youn Seo, Jeong-Yun Park, Sung-Jun Lee, Harim Lee, Yong-Hwan Department of Semiconductor System Engineering Kumoh National Institute of Technology Korea Republic of School of Electronic Engineering Kumoh National Institute of Technology Korea Republic of

In this paper, we introduce the design of an artificial intelligence (AI) hardware accelerator for a convolutional neural network (CNN) for classification tasks. The architecture of the AI hardware for a CNN model consists of two modules: a striding-padding (SP) module and a convolution layer module. For the classification task, fully-connected layers should follow the CNN model, and hence we also briefly explain the architecture of the fully-connected layer. In the architecture of a CNN model, the SP module manages how to access memories including feature maps and weights by considering the striding and padding operation of a convolution layer, and the convolution layer conducts convolution operation with feature maps and weights that are read from the memories. To verify the AI hardware design through RTL simulation, CNN models are trained in software using a MNIST dataset and a Fashion MNIST dataset. Then, post-training quantization and quantization-aware training, which are representative quantization techniques, are utilized to extract hardware-appropriate weights from the trained CNN models. The quantized weights are loaded to memories, and the validation data from the dataset used to train CNN models are fed to the designed CNN hardware. Finally, the inference results of the hardware are compared with that of the software-trained CNN models, which confirms the correctness of the hardware design. © The Korean Institute of Electrical Engineers.

关键词： Deep learning

来源：评论

学校读者我要写书评

暂无评论

Anticipating the lifespan: Predicting the durability of an anode-supported solid oxide fuel cell short stack over 50,000 h

引用

Progress in Natural Science:Materials International 2024年第3期34卷 606-613页

作者： Muhammad Zubair Khan Amjad Hussain Seung-Bok Lee Tak-Hyoung Lim Rak-Hyun Song Department of Materials Science & Engineering Pak-Austria Fachhochschule: Institute of Applied Sciences and Technology Hydrogen Energy Research Division Korea Institute of Energy Research Department of Advanced Energy and System Engineering Korea University of Science and Technology (UST)

In the present study, the operational lifetime of a solid oxide fuel(SOFC) short stack is predicted by investigating the performance degradation of both the short stack and its cells throughout 1000 h at 800°C. The short stack and integral cell voltages are continuously measured during the long-term test, with electrochemical impedance spectroscopy(EIS) conducted every 200 h. The short stack voltage decreased rapidly for the initial 200–300 h and afterwards, it decreased at a slow rate due to the increase in the Ohmic and polarization resistances in the same manner. Scanning electron microscopy results show that there is no delamination or cracking among constituent layers of the short-stack cells. The single degradation effects of the Ni coarsening in the anode, cation migration and surface segregation in cathode and oxide scale growth in metallic interconnect mesh are successfully integrated into a comprehensive lifetime prediction model. The experimentally measured voltage degradation data of the short stack fits well with the developed mathematical model and allows the successful prediction of the lifetime up to 50,000 h.

关键词： SOFC Stack Degradation Impedance Lifetime

来源：评论

学校读者我要写书评

暂无评论

Oversampling-Enhanced Feature Fusion-Based Hybrid ViT-1DCNN Model for Ransomware Cyber Attack Detection

引用

Computer Modeling in engineering & Sciences 2025年第2期142卷 1667-1695页

作者： Muhammad Armghan Latif Zohaib Mushtaq Saifur Rahman Saad Arif Salim Nasar Faraj Mursal Muhammad Irfan Haris Aziz Department of Computer and Information System Cleveland State UniversityOhio44115USA Department of Electrical Electronics and Computer SystemsCollege of Engineering and TechnologyUniversity of SargodhaSargodha40100Pakistan Electrical Engineering Department College of EngineeringNajran UniversityNajran61441Saudi Arabia Department of Mechanical Engineering College of EngineeringKing Faisal UniversityAl Ahsa31982Saudi Arabia Department of Mechanical Industrial and Energy System EngineeringUniversity of SargodhaSargodha40100Pakistan

Ransomware attacks pose a significant threat to critical infrastructures,demanding robust detection *** study introduces a hybrid model that combines vision transformer(ViT)and one-dimensional convolutional neural network(1DCNN)architectures to enhance ransomware detection *** common challenges in ransomware detection,particularly dataset class imbalance,the synthetic minority oversampling technique(SMOTE)is employed to generate synthetic samples for minority class,thereby improving detection *** integration of ViT and 1DCNN through feature fusion enables the model to capture both global contextual and local sequential features,resulting in comprehensive ransomware *** on the UNSW-NB15 dataset,the proposed ViT-1DCNN model achieved 98%detection accuracy with precision,recall,and F1-score metrics surpassing conventional *** approach not only reduces false positives and negatives but also offers scalability and robustness for real-world cybersecurity *** results demonstrate the model’s potential as an effective tool for proactive ransomware detection,especially in environments where evolving threats require adaptable and high-accuracy solutions.

关键词： Ransomware attacks cybersecurity vision transformer convolutional neural network feature fusion encryption threat detection

来源：评论

学校读者我要写书评

暂无评论

Controllability of heterogeneous networked sampled-data systems

引用

Science China(Information Sciences) 2025年第3期68卷 302-318页

作者： Zixuan YANG Lin WANG Xiaofan WANG Guanrong CHEN Department of Automation Shanghai Jiao Tong University Key Laboratory of System Control and Information Processing Ministry of Education of China School of Mechatronic Engineering and Automation Shanghai University School of Electrical and Electronic Engineering Shanghai Institute of Technology Department of Electrical Engineering City University of Hong Kong

This study investigates the controllability of a general heterogeneous networked sampled-data system(HNSS) consisting of nonidentical node systems, where the inner coupling between any pair of nodes can be described by a unique *** signals on control and transmission channels are sampled and held by zero-order holders, and the control sampling period of each node can be different. Necessary and sufficient controllability conditions are developed for the general HNSS, using the Smith normal form and matrix equations, respectively. The HNSS in specific topology or dynamic settings is discussed subsequently with easier-to-verify conditions derived. These heterogeneous factors have been determined to independently or jointly affect the controllability of networked sampled-data systems. Notably, heterogeneous sampling periods have the potential to enhance the overall controllability, but not for systems with some special dynamics. When the node dynamics are heterogeneous,the overall system can be controllable even if it is topologically uncontrollable. In addition, in several typical heterogeneous sampled-data multi-agent systems, pathological sampling of single-node systems will necessarily cause overall uncontrollability.

关键词： network controllability sampled-data system heterogeneous system inner coupling multi-agent system

来源：评论

学校读者我要写书评

暂无评论

Fuzzy Adaptive State Estimation of Distributed Drive Electric Vehicles with Random Missing Measurements and Unknown Process Noise

引用

Chinese Journal of Mechanical engineering 2024年第5期37卷 543-554页

作者： Zhiguo Zhang Guodong Yin Chao Huang Jingyu Hu Xing Xu Chengyue Jiang Yan Wang School of Mechanical Engineering Southeast UniversityNanjing 211189China Department of Industrial and System Engineering The Hong Kong Polytechnical UniversityHong KongChina Department of Automotive Engineering Research Institute Jiangsu UniversityZhenjiang 225009China College of Vehicle Engineering Chongqing University of TechnologyChongqing 400054China

Accurate estimation of sideslip angle and vehicle velocity is crucial for effective control of distributed drive electric ***,as these states are not directly measured,Kalman-based approaches utilizing in-vehicle sensors have been developed to estimate ***,existing methods tend to ignore the impact of data loss on estimation ***,the process noise,which changes dynamically due to varying driving conditions,is not adequately *** response to these constraints,we propose a novel method called the fuzzy adaptive fault-tolerant extended Kalman filter(FAFTEKF).Initially,a fault-tolerant EKF is devised to handle missing ***,a fuzzy logic system that dynamically updates the process noise matrix,is built to improve estimation accuracy under different driving *** experimental results validate the superiority of the FAFTEKF over the traditional EKF across various scenarios with different degrees of data loss.

关键词： Distributed drive electric vehicles State estimation Fault-tolerant EKF Fuzzy logic system

来源：评论

学校读者我要写书评

暂无评论

Observation and verification of surface electrical explosion driven by radial-distributed pulsed current in laboratory lightning strike test

引用

High Voltage 2024年第1期9卷 137-149页

作者： Ruoyu Han Yuchen Cao Yakun Liu Xi Chen Ting Li Chen Li Pengfei Li State Key Laboratory of Explosion Science and Technology Beijing Institute of TechnologyBeijingChina School of Mechatronical Engineering Beijing Institute of TechnologyBeijingChina Department of Electrical Engineering Shanghai Jiaotong UniversityShanghaiChina Beijing Institute of Astronautical System Engineering BeijingChina Aerospace Research Institute of Materials and Processing Technology BeijingChina

The laboratory lightning test is essential for assessing the effectiveness of lightning strike protection(LSP).Particularly,direct lightning strike damage can be performed with pulsed current injection into the *** paper focuses on the dynamic process of arc plasma and shock wave behaviour in the vicinity of the‘strike’point.A rod-plate discharge load is built for testing aluminium and coated plate under 40-kA-level pulsed *** visualisation of the luminous discharge plasma and its flow field via high-speed photography(from different angles)is meticulously designed and implemented,synchronised with electro-physical *** results indicate some new mechanisms for lightning strike damage,apart from the impulse heat loading from the thermal *** transient current injection through the arc root concentrates on a thin skin layer(skin-depth effect),with the radial-attenuated current density,driving asynchronously electrical explosions on the plate *** inhomogeneous Joule heating of the plate leads to outwardly propagating phase transition and shock wave along the conductive *** addition,the electro-thermal instability is observed and regarded as the seed of irregular erosion *** information reveals two different plasma states of main discharge arc channel and adjacent surface electrical *** correspondence of the physical mechanism of electrical explosion and optical radiation is *** images for different regions depict erosion characteristics and summarise influencing factors,further confirming the mechanism *** research clarifies the role of skin-depth effect in transaction arc erosion for electrode,complements the electrical explosion theory with unevenly distributed current and helps optimise strategies of LSP.

关键词： electrical lightning radial

来源：评论

学校读者我要写书评

暂无评论

Highly dense and ultrathin gadolinia-doped ceria interlayer for enhanced performance of large-area anode-supported solid oxide fuel cell

引用

Progress in Natural Science:Materials International 2023年第5期33卷 733-741页

作者： Sang Woo Shin Amjad Hussain Muhammad Zubair Khan Dong Woo Joh Tak-Hyung Lim Seung-Bok Lee Jong-Eun Hong Joon Hyung Shim Rak-Hyun Song Hydrogen Energy Research Division Korea Institute of Energy Research Department of Mechanical Engineering Korea University Department of Advanced Energy and System Engineering Korea University of Science and Technology(UST) Department of Materials Science & Engineering Pak-Austria Fachhochschule: Institute of Applied Sciences and Technology

The fabrication of a highly dense and ultrathin Gd doped ceria(GDC) interlayer on a large-area, porous electrolyte-coated planar anode support using vacuum slurry coating is studied for application in solid oxide fuel cells(SOFCs). The electrolyte-coated planar anode support body is fabricated by tape casting and pre-sintering is performed at 1100, 1150, and 1200°C. The GDC interlayer is coated onto the pre-sintered electrolyte-coated anode support using a vacuum slurry coating process with different immersion times(30–60 s). A highly dense and ultrathin GDC interlayer with a thickness of 1.48 μm is produced after co-sintering at 1400°C. The 1200°C pre-sintered electrolyte-coated anode support showed a thin and highly dense GDC interlayer which is attributed to a lower shrinkage difference between the coated GDC interlayer and pre-sintered body. The electrochemical performance of a vacuum slurry-coated GDC interlayer cell is 38 % higher compared to that of the screen-printed GDC-coated cell at 700°C due to the greatly reduced Ohmic and electrode polarization resistances. The considerable improvement in the performance is attributed to the ultrahigh GDC density which has prevented the secondary phase formation between the cathode and electrolyte. This study presents the economical, scalable, and reproducible process for the fabrication of highly robust GDC interlayer for the large-area planar anode-supported SOFCs.

关键词： Solid oxide fuel cell Vacuum slurry coating GDC interlayer Sintering Performance

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：