检索结果-内蒙古大学图书馆

60th Annual Allerton Conference on Communication, Control, and Computing, Allerton 2024

作者： Ghiasvand, Sajjad Reisizadeh, Amirhossein Alizadeh, Mahnoosh Pedarsani, Ramtin University of California Department of Electrical and Computer Engineering Santa BarbaraCA93106 United States Massachusetts Institute of Technology Laboratory for Information and Decision Systems CambridgeMA02139 United States

ISBN: (纸本)9798331541033

As distributed learning applications like Federated Learning, the Internet of Things (IoT), and Edge Computing expand, addressing their limitations becomes crucial. We approach decentralized learning across a network of communicating clients or nodes, focusing on two primary challenges: data heterogeneity and adversarial robustness. To address these, we introduce a decentralized minimax optimization method incorporating two key components: local updates and gradient tracking. Minimax optimization serves as a fundamental tool for adversarial training, ensuring robustness. Local updates are vital in Federated Learning (FL) to alleviate the communication bottleneck, while gradient tracking is necessary to demonstrate convergence amid data heterogeneity. Our analysis of the proposed algorithm, Dec-Fed Track, in nonconvex-strongly-concave minimax optimization demonstrates its convergence to a stationary point. Additionally, numerical experiments support our theoretical results. © 2024 IEEE.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

Camera Lidar Fusion for Unmanned Aerial Vehicle Detection

Camera Lidar Fusion for Unmanned Aerial Vehicle Detection

引用

Iberian robotics Conference (ROBOT)

作者： Cosimo Leuci Sedat Dogru Lino Marques Department of Electronics and Telecommunications (DET) Polytechnic University of Turin Turin Italy Department of Electrical and Computer Engineering Institute of Systems and Robotics University of Coimbra Coimbra Portugal

ISBN: (数字)9798350376364

ISBN: (纸本)9798350376371

Due to their low-cost and ease of use, cameras are very popular for drone detection. Thanks to the feature reach images obtained from the cameras, it is possible to classify drones with a relatively high success ratio. However, cameras lack the means for accurate localization, allowing only a rough measure of the orientation of the target UAV. In order to better localize the UAV, in this study we fuse camera and lidar measurements, localizing UAVs accurately. This fusion was implemented using a Kalman filter, which enables integration of both sensor data, improving detection and tracking accuracy

关键词： Location awareness YOLO Laser radar Accuracy Robot vision systems Cameras Autonomous aerial vehicles Kalman filters Vehicle dynamics Drones

来源：评论

学校读者我要写书评

暂无评论

An Efficient Data Collection using Whale Optimization Algorithm (WOA) in Wireless Sensor Network 2

An Efficient Data Collection using Whale Optimization Algori...

引用

2nd IEEE Global Conference on Communications and Information Technologies, GCCIT 2024

作者： Pasupathi, Udhayakumar Gopi Saminathan, A. Manikandan, A. Subhashini, M. Gowrishankar, C. Poornappriya, T.S. Department of Computer Science Tamilnadu India Npr College of Engineering and Technology Dindigul India Srm Institute of Science and Technology Chennai India Department of Computer Science Trichy India Kongunadu College of Engineering and Technology Department of Electrical and Electronics Engineering Tiruchirappalli621215 India Aspire Systems Pvt Ltd Karnataka 560037 India

ISBN: (纸本)9798350388916

Wireless sensor networks (WSNs) are an innovative technology that can be used in critical situations such as battlefields as well as commercial applications including buildings, traffic monitoring, smart homes, habitat monitoring, and more. A wireless sensor network (WSN) comprises small nodes capable of wireless sensing, processing, and communication. One of the most pressing concerns plaguing WSNs today is security. This paper suggests and investigates an energy-efficient cluster head (CH) selection approach using the Whale Optimization Algorithm (WOA) (called WOA-Clustering (WOA-C)) and the Geographic and Energy Aware Routing Algorithm (GEAR). As a result, the suggested method simplifies the decision-making process of energy-aware CHs by utilizing the fitness function that assesses node residual energy and the total energy of neighbouring nodes. Compared to the GWO-C, GABEEC, and PSOMSB protocols, the proposed WOA-GEAR protocol consumes 28%, 22%, and 7.5% less power. The suggested algorithm's network lifetime, power efficiency, buffer occupancy, throughput, and end-to-end latency were evaluated. © 2024 IEEE.

关键词： Decision making

来源：评论

学校读者我要写书评

暂无评论

A SiC-Based 243 kW 60 kW/L Electric Vehicle Traction Inverter

A SiC-Based 243 kW 60 kW/L Electric Vehicle Traction Inverte...

引用

2024 IEEE Energy Conversion Congress and Exposition, ECCE 2024

作者： Wu, Chih-Chiang Rout, Uma Sankar Chou, Hsin-Ping Hu, Jwu-Sheng Industrial Technology Research Institute Mechanical and Mechatronics Systems Research Laboratories Hsinchu Taiwan National Yang Ming Chiao Tung University Department of Electrical Engineering & Computer Science Hsinchu Taiwan

ISBN: (纸本)9798350376067

This paper investigates the evolving landscape of EV traction inverters, addressing key technical goals outlined by US Drive and analyzing parameters such as voltage, power density, and efficiency. Through innovative methodologies, a high-efficiency SiC-based traction inverter is developed to overcome challenges like spatial constraints. Leveraging international benchmarks, a cutting-edge SiC-based 800V 243 kW traction inverter with a power density of 60 kW/L is presented, poised to surpass US Drive's 2025 targets. This study emphasizes SiC technology's role in advancing EV propulsion systems towards sustainable and efficient transportation solutions. © 2024 IEEE.

关键词： power density silicon carbide high-efficiency sustainable transportation traction inverter

来源：评论

学校读者我要写书评

暂无评论

Energy Management System for a Smart Green Nanogrid feeding a Research Laboratory with Autonomous Mobile Robots

引用

IFAC-PapersOnLine 2024年第2期58卷 156-161页

作者： Matteo Fresia Lorenzo De Simone Ilaria Traverso Stefano Bracco Silvia Siri Department of Electrical Electronic Telecommunication Engineering and Naval Architecture University of Genoa 16145 Genova Italy Department of Computer Science Bioengineering Robotics and Systems Engineering University of Genoa 16145 Genova Italy

This paper proposes a mixed-integer linear programming optimization model used to define an energy management system tailored for nanogrids in buildings, integrating renewable energy sources, battery energy storage systems and task-executing autonomous mobile robots. Focused on a nanogrid to be realised at the Savona Campus of the University of Genoa, the energy management system optimizes power flows and robot task scheduling in order to minimize the operating costs, the curtailment of the photovoltaic source and the number of unperformed tasks. Its novelty lies in combining energy and task planning constraints, offering significant potential for sustainable building energy management.

关键词： Energy management system Optimal operation Nanogrids Prosumer building Autonomous mobile robots Renewable energy sources Battery energy storage system

来源：评论

学校读者我要写书评

暂无评论

PrivShape: Extracting Shapes in Time Series under User-Level Local Differential Privacy 40

PrivShape: Extracting Shapes in Time Series under User-Level...

引用

40th IEEE International Conference on Data engineering, ICDE 2024

作者： Mao, Yulian Ye, Qingqing Hu, Haibo Wang, Qi Huang, Kai Southern University of Science and Technology Department of Computer Science and Engineering China The Hong Kong Polytechnic University Department of Electrical and Electronic Engineering Hong Kong Research Institute of Trustworthy Autonomous Systems Southern University of Science and Technology China School of Computer Science and Engineering Macau University of Science and Technology China

ISBN: (纸本)9798350317152

Time series have numerous applications in finance, healthcare, IoT, and smart city. In many of these applications, time series typically contain personal data, so privacy infringement may occur if they are released directly to the public. Recently, local differential privacy (LDP) has emerged as the state-of-the-art approach to protecting data privacy. However, existing works on LDP-based collections cannot preserve the shape of time series. A recent work, PatternLDP, attempts to address this problem, but it can only protect a finite group of elements in a time series due to ?-event level privacy guarantee. In this paper, we propose PrivShape, a trie-based mechanism under user-level LDP to protect all elements. PrivShape first transforms a time series to reduce its length, and then adopts trie-expansion and two-level refinement to improve utility. By extensive experiments on real-world datasets, we demonstrate that PrivShape outperforms PatternLDP when adapted for offline use, and can effectively extract frequent shapes. © 2024 IEEE.

关键词： Time series

来源：评论

学校读者我要写书评

暂无评论

Optimized Rewinding of Oversized Induction Motors

Optimized Rewinding of Oversized Induction Motors

引用

IEEE International Conference on Electric Machines and Drives (IEMDC)

作者： Fernando J. T. E. Ferreira Dept. of Electrical and Computer Engineering Institute of Systems and Robotics University of Coimbra Coimbra Portugal

When an oversized and/or faulty three-phase squirrel-cage induction motor (SCIM) needs to be rewound, there is an opportunity to redesign its stator winding as a function of its actual load cycle, aiming at improving its efficiency, power factor, and reliability. This paper proposes a systematic methodology to optimize the stator winding of in-operation, line-operated, fixed-speed SCIMs, with two goals, namely, (i) optimization of the fundamental component of magnetizing flux to the actual load cycle, reducing the original core saturation level and nominal power of the motor, and (ii) optimization of the pitch and turns of the coils to minimize the winding copper losses and the airgap magnetomotive force harmonic content. Hybrid computing approaches using the first principles and nondeterministic methods can be used for winding optimization. The proposed methodology has been applied to three oversized 5.5 kW, 4-pole SCIMs, driving industrial ceramic paste mixers, and led to a 30-45% reduction in the line current, a 10-25% reduction in the input active power, and a 20-35% increase in the power factor, enabling an upgrade of the efficiency class from "IE0"/IE1 to IE3/IE4. The decrease of the motor line current reduces the Joule losses in the upstream cables and transformers, freeing power transmission capacity in the power supply network. The proposed strategy is a cost-effective alternative to the replacement of old oversized motors with new well-sized motors, avoiding mechanical readaptation of the fixing and coupling systems and allowing for higher efficiency and power factor levels than those of new well-sized motors. It also fosters copper recycling, the electric motor circular economy, and the local economic activity associated with the motor repair/rewinding sector.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Use of Virtual Reality to Facilitate Engineer Training in the Aerospace Industry

引用

Machine Graphics and Vision 2023年第2期32卷 19-44页

作者： Paszkiewicz, Andrzej Salach, Mateusz Wydrzyński, Dawid Woźniak, Joanna Budzik, Grzegorz Bolanowski, Marek Ganzha, Maria Paprzycki, Marcin Cierpicki, Norbert Department of Complex Systems Rzeszow University of Technology Rzeszów Poland Department of Machine Design Rzeszow University of Technology Rzeszów Poland Department of Management Systems and Logistics Rzeszow University of Technology Rzeszów Poland Systems Research Institute Polish Academy of Sciences Warszawa Poland The Faculty of Electrical and Computer Engineering Rzeszow University of Technology Rzeszów Poland

This work concerns automation of the training process, using modern information technologies, including virtual reality (VR). The starting point is an observation that automotive and aerospace industries require effective methods of preparation of engineering personnel. In this context, the technological process of preparing operations of a CNC numerical machine has been extracted. On this basis, a dedicated virtual reality environment, simulating manufacturing of a selected aircraft landing gear component, was created. For a comprehensive analysis of the pros and cons of the proposed approach, four forms of training, involving a physical CNC machine, a physical simulator, a software simulator, and the developed VR environment were instantiated. The features of each training form were analysed in terms of their potential for industrial applications. A survey, using the Net Promoter Score method, was also conducted among a target group of engineers, regarding the potential of use of each training form. As a result, the advantages and disadvantages of all four training forms were captured. They can be used as criteria for selecting the most effective training form. © 2023 Faculty of Applied Informatics and Mathematics - WZIM, Warsaw University of Life Sciences - SGGW. All rights reserved.

关键词： Aerospace industry

来源：评论

学校读者我要写书评

暂无评论

WildGaussians: 3D Gaussian Splatting in the Wild 38

WildGaussians: 3D Gaussian Splatting in the Wild

引用

38th Conference on Neural Information Processing systems, NeurIPS 2024

作者： Kulhanek, Jonas Peng, Songyou Kukelova, Zuzana Pollefeys, Marc Sattler, Torsten Czech Institute of Informatics Robotics and Cybernetics Czech Technical University in Prague Czech Republic Faculty of Electrical Engineering Czech Technical University in Prague Czech Republic Department of Computer Science ETH Zurich Switzerland Visual Recognition Group Faculty of Electrical Engineering Czech Technical University in Prague Czech Republic

While the field of 3D scene reconstruction is dominated by NeRFs due to their photorealistic quality, 3D Gaussian Splatting (3DGS) has recently emerged, offering similar quality with real-time rendering speeds. However, both methods primarily excel with well-controlled 3D scenes, while in-the-wild data - characterized by occlusions, dynamic objects, and varying illumination - remains challenging. NeRFs can adapt to such conditions easily through per-image embedding vectors, but 3DGS struggles due to its explicit representation and lack of shared parameters. To address this, we introduce WildGaussians, a novel approach to handle occlusions and appearance changes with 3DGS. By leveraging robust DINO features and integrating an appearance modeling module within 3DGS, our method achieves state-of-the-art results. We demonstrate that WildGaussians matches the real-time rendering speed of 3DGS while surpassing both 3DGS and NeRF baselines in handling in-the-wild data, all within a simple architectural framework. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Exploring Continuous Awareness Modelling for Improving Worker Safety and Trust*

Exploring Continuous Awareness Modelling for Improving Worke...

引用

IEEE International Workshop on Robot and Human Communication (ROMAN)

作者： Andrey Solovov Mohammad Zarei Vishal Gautam Bruno Ferreira Gustavo Assunção Antonio Marin-Hernandez Paulo Menezes Department of Electrical and Computer Engineering University of Coimbra Coimbra Portugal Institute of Systems and Robotics University of Coimbra Coimbra Portugal Universidad Veracruzana Xalapa Mexico

ISBN: (数字)9798350375022

ISBN: (纸本)9798350375039

The Industry 5.0 Human-Robot Interaction (HRI) goals have the potential for significantly reducing production costs, but require factory workers to trust that robots will not cause them physical harm when sharing the workspace in order to be achieved. To this end, we propose a model that adjusts the robot’s actions to an estimate of the workers’ awareness of its behavior. Its premise is that mutual awareness should be continuous, and if no recent visual contact has been established, then the robot should adapt its range of motion and speed in order to minimize the intersection between the robot’s space and the worker’s space. On the other hand, if the worker steadily establishes visual contact from times to times, the robot operates under a normal or a faster configuration. After a period of habituation, this awareness model can be further tweaked to continuously improve production times, while improving worker’s safety and trust. We present preliminary results of a pilot study with 32 participants, to asses if the existence of this awareness model impacts the outcomes of the combined task. The obtained tepidly optimistic results are the key to build upon in future designs.

关键词： Visualization Adaptation models Costs Service robots Human-robot interaction Production facilities Safety Motion control Fifth Industrial Revolution

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：