检索结果-内蒙古大学图书馆

arXiv 2021年

作者： Narang, Adhyyan Muthukumar, Vidya Sahai, Anant Electrical and Computer Engineering University of Washington Electrical and Computer Engineering and Industrial and Systems Engineering Georgia Institute of Technology Electrical Engineering and Computer Sciences UC Berkeley United States

State-of-the-art deep learning classifiers are heavily overparameterized with respect to the amount of training examples and observed to generalize well on "clean" data, but be highly susceptible to infinitesmal adversarial perturbations. In this paper, we identify an overparameterized linear ensemble, that uses the "lifted" Fourier feature map, that demonstrates both of these behaviors 1. The input is one-dimensional, and the adversary is only allowed to perturb these inputs and not the non-linear features directly. We find that the learned model is susceptible to adversaries in an intermediate regime where classification generalizes but regression does not. Notably, the susceptibility arises despite the absence of model misspecification or label noise, which are commonly cited reasons for adversarial-susceptibility. These results are extended theoretically to a random-Fourier-sum setup that exhibits double-descent behavior. In both feature-setups, the adversarial vulnerability arises because of a phenomenon we term spatial localization: the predictions of the learned model are markedly more sensitive in the vicinity of training points than elsewhere. This sensitivity is a consequence of feature lifting and is reminiscent of Gibb’s and Runge’s phenomena from signal processing and functional analysis. Despite the adversarial susceptibility, we find that classification with these features can be easier than the more commonly studied "independent feature" models. Copyright © 2021, The Authors. 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

Effect of Cell-to-Cell Thermal Imbalance and Cooling Strategy on Electric Vehicle Battery Performance and Longevity

Effect of Cell-to-Cell Thermal Imbalance and Cooling Strateg...

引用

Intersociety Conference on Thermal and Thermomechanical Phenomena in Electronic systems (ITHERM)

作者： Camilo Escobar Zhe Gong Carlos Da Silva Olivier Trescases Cristina H. Amon Department of Mechanical and Industrial Engineering University of Toronto Toronto Canada The Edward S. Rogers Sr. Department of Electrical and Computer Engineering University of Toronto Toronto Canada Departments of Mechanical and Industrial Engineering Chemical Engineering and Applied Chemistry University of Toronto Toronto Canada

ISBN: (数字)9781665485036

ISBN: (纸本)9781665485043

This work develops a reduced-order numerical model of a custom-built commuter electric vehicle (EV) to study the lifetime performance of EV battery packs and battery thermal management systems (BTMS). The model uses experimental battery and BTMS data collected from a commuter EV and applies drive cycles corresponding to typical highway and city driving conditions. The main advantage of this numerical modeling approach is its ability to simulate large timescales, spanning years of vehicle operation. Cell level degradation is captured, allowing for the study of battery pack longevity under a variety of temperature profiles generated by various BTMS strategies with series and parallel indirect liquid cooling configurations. Monte Carlo simulations are also used to estimate the variation in BTMS performance caused by beginning-of-life (BOL) variations and cell-to-cell thermal imbalance/spreading in the battery cells. The proposed modeling approach was demonstrated to be an effective tool in studying long timescale BTMS performance, and tradeoffs between BTMS energy consumption and pack energy retention for the case-study commuter vehicle. A 7% reduction in the mean maximum pack temperature and a 10% reduction in the mean lifetime BTMS energy consumption were achieved by tuning BTMS control parameter thresholds. However, the variability in pack energy retention greatly increased, highlighting the need to consider BOL variations and cell spreading in BTMS modeling and design.

关键词： Degradation Thermal management of electronics Energy consumption Microprocessors computer architecture Thermal management Electric vehicles

来源：评论

学校读者我要写书评

暂无评论

Operational strategies for a photovoltaic-thermal system integrated with an electric air-to-water heat pump

Operational strategies for a photovoltaic-thermal system int...

引用

International Conference on Energy Transition in the Mediterranean Area (SyNERGY MED)

作者： Giuseppe Russo Andreas V. Olympios Chandan Pandey Annamaria Buonomano Christos N. Markides Department of Chemical Engineering Clean Energy Processes (CEP) Laboratory and Sargent Centre for Process Systems Engineering Imperial College London UK Department of Industrial Engineering University of Naples Federico II Naples Italy Department of Electrical and Computer Engineering PV Technology Laboratory University of Cyprus Nicosia Cyprus PHAETHON Centre of Excellence for Intelligent Efficient and Sustainable Energy Solutions Nicosia Cyprus

ISBN: (数字)9798350375923

ISBN: (纸本)9798350375930

In this study, operational strategies are investigated for an air-based photovoltaic-thermal (PV-T) system integrated with a domestic air-to-water heat pump. The PV-T system is used to simultaneously generate electricity and heat ambient air, which serves as the heat source for the heat pump. By utilising a variable-speed fan to regulate airflow, the PV-T system outlet temperature is adjusted, and the focus is on how different outlet temperatures impact the PVT system electrical and thermal efficiency, heat pump performance, and overall electricity requirements for heating. The performance of the integrated system is assessed for a case study of a household application in north Italy. The optimal PV-T outlet temperature is found to be between 30 – 35 °C, resulting in only 5% less electricity production than an equivalent PV system with equal nominal electricity capacity, while reducing overall electricity required for heating by over 15% compared to PV-integrated heat pumps.

关键词： Photovoltaic systems Fans Temperature Heat pumps Electricity Buildings Water heating Production Temperature control Resistance heating

来源：评论

学校读者我要写书评

暂无评论

Open source solution for CNC milling machine monitoring and comparison with Matlab simulation in real time

Open source solution for CNC milling machine monitoring and ...

引用

International Conference on Modern electrical and Energy systems (MEES)

作者： Branislav Fecko Nikitina Aliona Evgeniia Burdilna Department of Theoretical and Industrial Electrical Engineering Technical University of Košice Košice Slovak Republic Department of Automation and Computer Systems Kremenchuk Mykhailo Ostrohradskyi National University Kremenchuk Ukraine

This research paper focuses on the use of a thermal model of a CNC machine monitoring system that has been developed using open-source tools. By integrating these tools, predictive maintenance capabilities can be developed, leading to increased machine reliability. The methodology involves the implementation of a thermal model designed with Simulink software. The thermal model has been implemented into a developed monitoring system based on open-source tools, which allows access to real-time data of the CNC machine. The thermal model’s correctness has been verified with the machine operation using the data obtained from the CNC machine monitoring system. The use of open-source tools provides flexibility and adaptability in tailoring the thermal model to specific requirements and gaining key insights into the thermal behavior of the machine. The final section of the publication presents a comparison of simulated and measured outputs.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Exploring the effect of clustering algorithms on sample average approximation

Exploring the effect of clustering algorithms on sample aver...

引用

IISE Annual Conference and Expo 2021

作者： Jacobson, Daniel Hassan, Menna Dong, Zhijie Sasha Grado Department of Industrial and Systems Engineering Virginia Tech Falls ChurchVA United States School of Electrical and Computer Engineering Purdue University LafayetteIN United States Ingram School of Engineering Texas State University San MarcosTX United States

ISBN: (纸本)9781713838470

Stochastic Programming (SP) is used in disaster management, supply chain design, and other complex problems. Many of the real-world problems that SP is applied to produce large-size models. It is important but challenging that they are optimized quickly and efficiently. Existing optimization algorithms are limited in capability of solving these larger problems. Sample Average Approximation (SAA) method is a common approach for solving large scale SP problems by using the Monte Carlo simulation. This paper focuses on applying clustering algorithms to the data before the random sample is selected for the SAA algorithm. Once clustered, a sample is randomly selected from each of the clusters instead of from the entire dataset. This project looks to analyze five clustering techniques compared to each other and compared to the original SAA algorithm in order to see if clustering improves both the speed and the optimal solution of the SAA method for solving stochastic optimization problems. © 2021 IISE Annual Conference and Expo 2021. All rights reserved.

关键词： Machine learning

来源：评论

学校读者我要写书评

暂无评论

Faster Convergence of Local SGD for Over-Parameterized Models

arXiv

引用

arXiv 2022年

作者： Qin, Tiancheng Rasoul Etesami, S. Uribe, Cesár A. Department of Industrial and Systems Engineering Coordinated Science Laboratory University of Illinois at Urbana-Champaign United States Department of Electrical and Computer Engineering Rice University United States

Modern machine learning architectures are often highly expressive. They are usually overparameterized and can interpolate the data by driving the empirical loss close to zero. We analyze the convergence of Local SGD (or FedAvg) for such over-parameterized models in the heterogeneous data setting and improve upon the existing literature by establishing the following convergence rates. For general convex loss functions, we establish an error bound of O(1/T) under a mild data similarity assumption and an error bound of O(K/T) otherwise, where K is the number of local steps and T is the total number of iterations. For non-convex loss functions we prove an _error bound of O(K/T). These bounds improve upon the best previous bound of O(1/√nT) in both cases, where n is the number of nodes, when no assumption on the model being over-parameterized is made. We complete our results by providing problem instances in which our established convergence rates are tight to a constant factor with a reasonably small stepsize. Finally, we validate our theoretical results by performing large-scale numerical experiments that reveal the convergence behavior of Local SGD for practical over-parameterized deep learning models, in which the O(1/T) convergence rate of Local SGD is clearly shown. Copyright © 2022, The Authors. All rights reserved.

关键词： Parameterization

来源：评论

学校读者我要写书评

暂无评论

Protection against Counterfeiting Attacks in 3D Printing by Streaming Signature-embedded Manufacturing Process Instructions 1

Protection against Counterfeiting Attacks in 3D Printing by ...

引用

1st ACM International Workshop on Additive Manufacturing (3D Printing) Security, AMSec 2021

作者： Tiwari, Akash Villasenor, Eduardo Jose Gupta, Nikhil Reddy, Narasimha Karri, Ramesh Bukkapatnam, Satish T. S. Industrial and Systems Engineering Texas A and M University College StationTX United States Mechanical and Aerospace Engineering New York University BrooklynNY United States Electrical and Computer Engineering Texas A and M University College StationTX United States

ISBN: (纸本)9781450384803

The emerging Manufacturing-as-a-Service (MaaS) paradigm democratizes manufacturing by connecting people and businesses with manufacturing requests to those with manufacturing resources, via a digital thread. However, the digital thread can be vulnerable to attacks such as counterfeiting, IP theft and sabotage. An approach based on embedding custom anti-counterfeiting signatures in the design files, on-the-fly, is presented to protect against counterfeiting attacks. An experimental study of the effect of geometric and dimensional variations of the signatures in printed components, as well as the effect of the variations on the read rates of the codes are reported. We conduct experiments using a polyjet printer with veroclear material to study the positional and dimensional variations introduced in the embedded signature and the resulting effect on the readability of the design. The results suggest that signatures with spherical units of size †0.25 mm embedded within 20 mm cubic components can be printed with location precision of ∼5% (about 0.1 mm), and dimensional deviation of the order of 0.01 mm. A statistical model is developed to show that these variations pose minimum interference on the readability of the signatures. Embedding of randomized signatures offers security by serving as anti-counterfeit marks in the final part, and makes it harder to reverse engineer and produce counterfeited parts. © 2021 ACM.

关键词： 3D printers

来源：评论

学校读者我要写书评

暂无评论

Next Generation Advanced Transceiver Technologies for 6G and Beyond

arXiv

引用

arXiv 2024年

作者： You, Changsheng Cai, Yunlong Liu, Yuanwei Di Renzo, Marco Duman, Tolga M. Yener, Aylin Swindlehurst, A. Lee Department of Electronic and Electrical Engineering Southern University of Science and Technology Shenzhen China College of Information Science and Electronic Engineering Zhejiang University Hangzhou310027 China School of Electronic Engineering and Computer Science Queen Mary University of London LondonE1 4NS United Kingdom Université Paris-Saclay CNRS CentraleSupélec Laboratoire des Signaux et Systèmes Gif-sur-Yvette91192 France Department of Electrical and Electronics Engineering Bilkent University Ankara06800 Turkey Departments of Electrical and Computer Engineering Computer Science and Engineering and Integrated Systems Engineering The Ohio State University OH United States Center for Pervasive Communications and Computing University of California IrvineCA92697 United States

To accommodate new applications such as extended reality, fully autonomous vehicular networks and the metaverse, next generation wireless networks are going to be subject to much more stringent performance requirements than the fifth-generation (5G) in terms of data rates, reliability, latency, and connectivity. It is thus necessary to develop next generation advanced transceiver (NGAT) technologies for efficient signal transmission and reception. In this tutorial, we explore the evolution of NGAT from three different perspectives. Specifically, we first provide an overview of new-field NGAT technology, which shifts from conventional far-field channel models to new near-field channel models. Then, three new-form NGAT technologies and their design challenges are presented, including reconfigurable intelligent surfaces, flexible antennas, and holographic multi-input multi-output (MIMO) systems. Subsequently, we discuss recent advances in semantic-aware NGAT technologies, which can utilize new metrics for advanced transceiver designs. Finally, we point out other promising transceiver technologies for future research. Copyright © 2024, The Authors. All rights reserved.

关键词： Near field communication

来源：评论

学校读者我要写书评

暂无评论

Sharp global convergence guarantees for iterative nonconvex optimization: A Gaussian process perspective

arXiv

引用

arXiv 2021年

作者： Chandrasekhery, Kabir Aladin Pananjady, Ashwin Thrampoulidis, Christos Department of Electrical Engineering Stanford University Schools of Industrial & Systems Engineering and Electrical & Computer Engineering Georgia Tech Department of Electrical & Computer Engineering University of British Columbia

We consider a general class of regression models with normally distributed covariates, and the associated nonconvex problem of fitting these models from data. We develop a general recipe for analyzing the convergence of iterative algorithms for this task from a random initialization. In particular, provided each iteration can be written as the solution to a convex optimization problem satisfying some natural conditions, we leverage Gaussian comparison theorems to derive a deterministic sequence that provides sharp upper and lower bounds on the error of the algorithm with sample-splitting. Crucially, this deterministic sequence accurately captures both the convergence rate of the algorithm and the eventual error floor in the finite-sample regime, and is distinct from the commonly used "population" sequence that results from taking the infinite-sample limit. We apply our general framework to derive several concrete consequences for parameter estimation in popular statistical models including phase retrieval and mixtures of regressions. Provided the sample size scales near-linearly in the dimension, we show sharp global convergence rates for both higher-order algorithms based on alternating updates and first-order algorithms based on subgradient descent. These corollaries, in turn, yield multiple consequences, including: (a) Proof that higher-order algorithms can converge significantly faster than their first-order counterparts (and sometimes super-linearly), even if the two share the same population update and (b) Intricacies in super-linear convergence behavior for higher-order algorithms, which can be nonstandard (e.g., with exponent 3/2) and sensitive to the noise level in the problem. We complement these results with extensive numerical experiments, which show excellent agreement with our theoretical predictions. Copyright © 2021, The Authors. All rights reserved.

关键词： Normal distribution

来源：评论

学校读者我要写书评

暂无评论

Analysis of Magnetic Field Simulations of Magnetoelastic Sensors Quasi-Statically using COMSOL Multiphysics and MATLAB Software

Analysis of Magnetic Field Simulations of Magnetoelastic Sen...

引用

International Conference on Modern electrical and Energy systems (MEES)

作者： Šimon Gans Ján Molnár Vitalii Naida Mariia Maliakova Department of Theoretical and Industrial Electrical Engineering Technical University of Košice Košice Slovakia Department of Theoretical and Electrical Engineering Technical University of Košice Košice Slovakia Department of Automation and Computer Systems Kremenchuk Mykhailo Ostrohradskyi National University Kremenchuk Ukraine Department of Electrical Engineering Kremenchuk Mykhailo Ostrohradskyi National University Kremenchuk Ukraine

This work describes a procedure in which a transient simulation of electromagnetic fields can be done when the electrical conductivity of the ferromagnetic medium can be effectively neglected transforming it into a quasi-static problem. A pressductor-type magnetoelastic sensor simulation model was created. The sensor’s output consists of a change of the induced voltage in the secondary coil due to applied compressive stress. Induced voltage is proportional to the time-derivative of the magnetic flux flowing through the coil and therefore is transient in nature. However, transient analyses tend to be computationally expensive and can yield noisy outputs when the model is not sufficiently densely discretized and appropriate tolerances are chosen. To overcome this, a quasi-stationary analysis of the same system is introduced and computed using the COMSOL Multiphysics FEM simulation software with further processing in the MATLAB environment, and the results are compared.

关键词：

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：