检索结果-内蒙古大学图书馆

2nd IEEE International Conference on Control, Electronics and computer Technology, ICCECT 2024

作者： Chen, Zhaoxuan Jia, Haixin Peng, Jianbo Liu, Huilin School of Computer Science and Engineering Northeastern University Shenyang China School of Mechanical and Electrical Engineering China University of Mining and Technology-Beijing Beijing China

ISBN: (纸本)9798350380958

In road surface crack segmentation algorithms, obtaining comprehensive contextual information is crucial. While many solutions use the Transformer architecture for global information, its computational complexity requires combining it with CNN for feature extraction, making multi-scale contextual information extraction challenging. Although UNet-based networks are popular for fine-grained features, their decoder introduces redundant contents. To address above-mentioned issues, The paper proposes DBAUNet, incorporating SVD-ASPP (Same Value Dense ASPP) for global information extraction and DBAM (Dual Branch Attention Mechanism) to enhance crack information while suppressing background information. SVD-ASPP prevents crack disconnections, ensuring more accurate segmentation results. © 2024 IEEE.

关键词： Information retrieval

来源：评论

学校读者我要写书评

暂无评论

Generating Assembly Instructions Using Reinforcement Learning in Combination with Large Language Models 22

Generating Assembly Instructions Using Reinforcement Learnin...

引用

22nd IEEE International Conference on Industrial Informatics, INDIN 2024

作者： Widulle, Niklas Meyer, Frederic Niggemann, Oliver Computer Science in Mechanical Engineering Helmut Schmidt University Hamburg Germany Industrial Engineering Laboratory OWL University of Applied Sciences and Arts Lemgo Germany

ISBN: (纸本)9798331527471

The efficiency of manual assembly can be significantly improved by utilizing assistance systems that display assembly instructions. However, generating and maintaining these instructions require substantial effort, especially for low-volume or highly customized products. If neglected, this can lead to outdated instructions, frustrated operators, and the abandonment of the assistance platform. Recent advancements in Large Language Models (LLMs) have made it possible to generate high-quality assembly instructions given the right input. However, relying solely on LLMs to plan the assembly process risks producing unfeasible assembly sequences due to potential hallucinations by the models. To address this, Reinforcement Learning (RL) can be used in simulations to plan the assembly process, imposing restrictions on impractical movements. We propose a framework that integrates RL and LLMs to generate practical and accurate assembly instructions. In our framework, RL is used within a simulation to generate feasible assembly sequences. These sequences are then transformed into detailed assembly instructions by an LLM. We evaluate our framework on real-world product assemblies, generating comprehensive assembly instructions from corresponding CAD files. Our results demonstrate the potential of combining RL and LLMs to automate the generation of assembly instructions, thereby overcoming the limitations of current assistance systems and enhancing the efficiency of manual assembly processes. © 2024 IEEE.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Research on Image Encryption Algorithm Based on Chaotic System 4

Research on Image Encryption Algorithm Based on Chaotic Syst...

引用

4th IEEE International Conference on Data science and computer Application, ICDSCA 2024

作者： Yihang, Wu Yanling, Sun Yingqi, Li School of Mechanical Engineering Shenyang Jianzhu University Shenyang China Department of Science Shenyang Jianzhu University Shenyang China Computer Science and Engineering Shenyang Jianzhu University Shenyang China

ISBN: (纸本)9798350368239

With the development of the internet, social software and media have become ubiquitous, making images increasingly important in daily life. The dissemination of a large number of images on social networks has made information security issues increasingly prominent, and image encryption technology has become increasingly important. This article mainly studies image encryption algorithms based on chaotic systems. Firstly, the background and history of image encryption were introduced, followed by an explanation of the basic concepts based on chaotic systems. Then, several chaos based image encryption schemes were comprehensively compared, and their advantages and disadvantages were analyzed. Subsequently, several common measures for evaluating the effectiveness of encryption schemes were introduced. Finally, the future development of image encryption algorithms based on chaotic systems is discussed. ©2024 IEEE.

关键词： Encryption algorithms

来源：评论

学校读者我要写书评

暂无评论

New Passive Damped-LCL-Filter Based Three-Phase Four-Leg Shunt Active Power Filter

New Passive Damped-LCL-Filter Based Three-Phase Four-Leg Shu...

引用

2024 IEEE Energy Conversion Congress and Exposition, ECCE 2024

作者： Wang, Houqing Zhao, Yue Cao, Hui Peyman, Darvish Han, Dongfang Ma, Zhuxuan Zhang, Jinlai Wang, Zhenqi University of Arkansas Dept. of Electrical Engineering and Computer Science FayettevilleAR United States Xinjiang University Dept. of Computer Science and Technology Urumqi China Changsha University of Science and Technology Dept. of Mechanical and Electrical Engineering Changsha China

ISBN: (纸本)9798350376067

To suppress the resonance in an LCL filter, the passive damping method is often favored over the active damping due to its simplicity and robustness. However, the passive damping suffers from decreasing LCL filter's effectiveness in suppressing high-frequency harmonics and introducing power losses. In this paper, a new passive damping method is proposed for the three-phase four-leg shunt active power filter (SAPF) system, by adding resistors between the neutral line and the connection point of the three-phase lines. The selection of the damping resistor is guided by criteria that aim to minimize power losses while ensuring good stability characteristics. The SAPF system with the proposed passive damping demonstrates excellent attenuation of switching frequency harmonics and significant mitigation of interphase switching interference, enhancing harmonic compensation for nonlinear loads. The mathematical model based on single-phase equivalent circuit and control strategy of three-phase four-leg SAPF are analyzed respectively. Simulations and experiments with imbalanced loads are constructed to validate the theoretical analysis. © 2024 IEEE.

关键词： Equivalent circuits

来源：评论

学校读者我要写书评

暂无评论

Onboard Near-Real-Time Detection of Harmful Algal Blooms Using a Convolutional Neural Network and Multispectral Imagery from a Satellite in Low Earth Orbit

Onboard Near-Real-Time Detection of Harmful Algal Blooms Usi...

引用

OCEANS 2024 - Halifax, OCEANS 2024

作者： Parsons, Jonathan D. Seto, Mae L. Dalhousie University Computer Science Dept. Halifax Canada Dalhousie University Mechanical Engineering Dept. Halifax Canada

ISBN: (纸本)9798331540081

Freshwater harmful algal blooms (HABs) pose significant ecological and public health risks worldwide. Detecting HABs soon after they form is critical to managing the damage they cause. While in-situ measurements are more accurate at detecting and measuring their toxicity levels, satellite imagery is more adept at capturing the spatial and temporal dynamics of these blooms over large geographic regions. Satellites can also more persistently monitor for HABs. In the past, empirical methods and machine learning methods have used multispectral satellite imagery to estimate HAB biomass. To build upon the current body of research, this paper investigates an approach to expedite HAB detection by utilizing a convolutional neural network (CNN) deployed onboard a CubeSat in low Earth orbit to detect HABs in near-real-time. The CNN is trained with multispectral imagery from the Sentinel-2 satellite constellation aggregated with in-situ cyanobacteria cell counts from the Seabass CAML dataset. The results successfully demonstrated the capability of a CNN to detect cyanobacterial blooms using multispectral imagery. After classifying HAB predictions into 5 severity classes, the best performing model achieved a RMSE of 1.33 between HAB severity levels. Training the CNN on 30m GSD imagery with RGB and red edge (B05) bands achieved a RMSE of 1.83 between HAB severity levels, which was inadequate for detecting HABs in small inland water bodies. Improved performance was observed with 10m ground sample distance (GSD) band combinations. The best performing networks utilized all of Sentinel-2's 10m and 20m spectral bands. © 2024 IEEE.

关键词： Satellite imagery

来源：评论

学校读者我要写书评

暂无评论

Securing Cloud-Encrypted Data:Detecting Ransomware-as-a-Service(RaaS)Attacks through Deep Learning Ensemble

引用

computers, Materials & Continua 2024年第4期79卷 857-873页

作者： Amardeep Singh Hamad Ali Abosaq Saad Arif Zohaib Mushtaq Muhammad Irfan Ghulam Abbas Arshad Ali Alanoud Al Mazroa School of Computer Information Sciences University of CumberlandsWilliamsburgKentucky40769USA Computer Science Department College of Computer Science and Information SystemsNajran UniversityNajran61441Saudi Arabia Department of Mechanical Engineering College of EngineeringKing Faisal UniversityAl-Ahsa31982Saudi Arabia Department of Electrical Electronics and Computer SystemsCollege of Engineering and TechnologyUniversity of SargodhaSargodha40100Pakistan Electrical Engineering Department College of EngineeringNajran UniversityNajran61441Saudi Arabia Department of Electrical Engineering The University of LahoreLahore54000Pakistan Faculty of Computer and Information Systems Islamic University of MadinahMadinah42351Saudi Arabia Department of Information Systems College of Computer and Information SciencesPrincess Nourah bint Abdulrahman UniversityP.O.Box 84428Riyadh11671Saudi Arabia

Data security assurance is crucial due to the increasing prevalence of cloud computing and its widespread use across different industries,especially in light of the growing number of cybersecurity threats.A major and everpresent threat is Ransomware-as-a-Service(RaaS)assaults,which enable even individuals with minimal technical knowledge to conduct ransomware *** study provides a new approach for RaaS attack detection which uses an ensemble of deep learning *** this purpose,the network intrusion detection dataset“UNSWNB15”from the Intelligent Security Group of the University of New South Wales,Australia is *** the initial phase,the rectified linear unit-,scaled exponential linear unit-,and exponential linear unit-based three separate Multi-Layer Perceptron(MLP)models are ***,using the combined predictive power of these three MLPs,the RansoDetect Fusion ensemble model is introduced in the suggested *** proposed ensemble technique outperforms previous studieswith impressive performance metrics results,including 98.79%accuracy and recall,98.85%precision,and 98.80%*** empirical results of this study validate the ensemble model’s ability to improve cybersecurity defenses by showing that it outperforms individual *** expanding the field of cybersecurity strategy,this research highlights the significance of combined deep learning models in strengthening intrusion detection systems against sophisticated cyber threats.

关键词： Cloud encryption RaaS ensemble threat detection deep learning cybersecurity

来源：评论

学校读者我要写书评

暂无评论

An Obstacle Avoiding Motion Planning Method for Manipulator using Progressive Dynamic Local Search 4

An Obstacle Avoiding Motion Planning Method for Manipulator ...

引用

4th International Symposium on Intelligent Robotics and Systems, ISoIRS 2024

作者： Li, Na Ouyang, Yun Deng, Jun Gao, Zhenguo Jiang, Kunliang Huaqiao University School of Mechanical Engineering and Automation Xiamen China Huaqiao University School of Computer Science and Technology Xiamen China

ISBN: (纸本)9798350364798

To enhance the efficiency of motion planning, we introduce a new angle profile searching method known as Progressive Dynamic Local Search (PDLS). Building upon a collision-free multi-segment path established by the RSPM-PS algorithm, PDLS methodically computes angle profiles from the initial to the target pose, focusing searches near each point's predecessor. It refines the path's smoothness by adding intermediate points and sets the local search radius through the Jacobian matrix, optimizing convergence. Our method also innovates in assessing angle profile quality, introducing a combined error function that judiciously weighs the importance of position and attitude accuracy. Furthermore, we mitigate sudden changes in joint acceleration by interpolating angle profiles using a quintic polynomial. Tests in MATLAB and on a JAKA Zu311 manipulator demonstrate PDLS's advantages: reduced run times, precise calculations, and improved convergence rates. © 2024 IEEE.

关键词： Jacobian matrices

来源：评论

学校读者我要写书评

暂无评论

Real-Time Threat Management using Deep Q-Learning and Mininet 3

Real-Time Threat Management using Deep Q-Learning and Minine...

引用

3rd International Conference on Intelligent Data Communication Technologies and Internet of Things, IDCIoT 2025

作者： Sivakumar, Amrutha Bhavya Reddy, G. Bhanot, Shreya Rajagopal, Shinu M. Prashanth, B.N. Department of Computer Science and Engineering Amrita School of Computing Amrita Vishwa Vidyapeetham Bengaluru India Department of Mechanical Engineering Amrita School of Engineering Amrita Vishwa Vidyapeetham Bengaluru India

ISBN: (纸本)9798331527549

The research combines Deep Q-Learning(DQN) with a Mininet-based network simulation and Scapy intrusions detection system (IDS) for malicious traffic prioritizing. The RL agent continuously learns to act based on real-time traffic data and builds an adaptive model that detects and prioritizes threats such as SYN Flood, UDP Flood, Slowloris, HTTP Flood etc. The RL agent then adjusts its policy via the seamless combination of network simulation and live traffic detection to make sure lossy channels representing high-severity attacks are dealt with first, lessening the damage. The system improves the identification of intrusions by stimulating the intrusion prioritization dynamically with continuous learning and feedback, increasing network security. © 2025 IEEE.

关键词： Network intrusion

来源：评论

学校读者我要写书评

暂无评论

Deadline-aware and energy efficient IoT task scheduling using fuzzy logic in fog computing

引用

Multimedia Tools and Applications 2025年第15期84卷 14359-14386页

作者： Thakur, Rahul Sikka, Geeta Bansal, Urvashi Giri, Jayant Mallik, Saurav Department of Computer Science and Engineering Dr. B R Ambedkar National Institute of Technology Punjab Jalandhar144011 India Department of Computer Science and Engineering National Institute of Technology Delhi Delhi110036 India Department of Mechanical Engineering Yeshwantrao Chavan College of Engineering Nagpur India Department of Environmental Health Harvard T H Chan School of Public Health BostonMA United States Saveetha University TN Chennai602105 India

Fog computing is an emerging paradigm that extends cloud computing (CC) by providing computation, communication, and storage services at the edge of a network, closer to end devices. It has gained significance due to the rapid development of IoT devices, which generate various types of tasks. Processing these tasks in the cloud can strain its infrastructure and lead to delays in time-sensitive requests. To address this limitation, fog computing (FC) concepts were introduced in 2012 by Cisco. FC is not meant to replace CC but rather to complement and extend its capabilities. One of the challenges in FC is efficiently assigning tasks to appropriate resources to minimize makespan, energy consumption (EC), and increase the number of deadline-satisfied tasks. In this work, the improvement of semi-greedy algorithm has been done by incorporating fuzzy logic (FL). By leveraging FL, the aim is to enhance the algorithm's decision-making process and make it more adaptive to varying conditions and uncertainties in the fog environment. The use of FL allows more nuanced and flexible task scheduling (TS) decisions based on fuzzy sets and fuzzy rules. The simulation experiments demonstrate that the proposed algorithm outperforms PSG (Priority-aware Semi-Greedy) and PSG-M (PSG with multistart), which were identified as the best scheduling algorithms (Algos) in the literature review. The algorithm exhibits better performance in terms of reducing makespan, EC, and increasing the percentage of deadline-satisfied tasks compared to PSG and PSG-M. The inclusion of FL further enhances the algorithm's effectiveness in handling complex scheduling scenarios in a FC environment. To evaluate the performance of the proposed algorithm, different simulation experiments have been conducted using a selected simulator after a systematic review of existing simulators. The experiments involved 300 and 500 random and static tasks, as well as 60 fog nodes in the fog environment. All simulations were impl

关键词： Fog computing

来源：评论

学校读者我要写书评

暂无评论

KB4RL: Towards a Knowledge Base for automatic creation of State and Action Spaces for Reinforcement Learning

KB4RL: Towards a Knowledge Base for automatic creation of St...

引用

Joint of the ESWC 2024 Workshops and Tutorials, ESWC-JP 2024

作者： Joualy, Lobna Demeester, Eric Tsiogkas, Nikolaos Department of Computer Science KU Leuven Diepenbeek3590 Belgium Department of Mechanical Engineering KU Leuven Diepenbeek3590 Belgium

Reinforcement Learning (RL) is a promising approach for creating adaptive solutions for robotic tasks that are difficult to design directly. Unlike traditional approaches that rely on designing explicit behaviors, RL allows agents to learn skills by interacting with their environment. Inspired by human learning processes, agents acquire knowledge through trial and error guided by the rewards gained from their actions through various experiences. A key step when solving planning problems with RL algorithms is the definition of state and action spaces. Typically, state and action spaces are manually designed in a domain-specific manner to ensure effective problem solving. In this paper, we present an innovative approach to define state and action spaces through a Knowledge Base (KB) encoding information about the environment and physics. We propose a structured and extensible form for knowledge description, and we show how to extend the action space through action learning. This allows the robot to reach its goal in those states where unexpected scenarios are encountered and where, therefore, the action space obtained through the KB is not sufficient. © 2024 Copyright for this paper by its authors.

关键词： Automatic action space Automatic state space Knowledge graphs Reinforcement learning

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：