检索结果-南通市图书馆

Accurate Distinction of Weak and Strong AC Grids for emerging Hierarchical-Infeed LCC-UHVDC Systems

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

CSee Journal of Power and energy Systems 2024年第2期10卷 772-777页

作者： Hao Xiao Xianzhong Duan Yi Zhang department of electrical and Computer engineering University of ManitobaWinnipeg R3T 2N2ManitobaCanada State Key Laboratory of Advanced electromagnetic engineering and technology Hubei Electric Power Security and High Efficiency Key LaboratorySchool of Electrical and Electronic EngineeringHuazhong University of Science and TechnologyWuhan 430074HubeiChina RtDS technologies Inc. Winnipeg R3T 2E1ManitobaCanada

Distinction of weak and strong AC grids for emerging hierarchical-infeed LCC-UHVDC systems is important for planning and operation departments. However, accuracy of earlier distinction methods is limited as t.ey were developed by empirical reasoning without rigorous t.eoretical analysis. Hence in this let.er, hierarchical-infeed interactive effective short-circuit ratio (HIeSCR) index is first used for st.ength evaluation of HIDC systems with complex inter-inverter interactions considered. Boundary HIeSCR (BHIeSCR) is also introduced in t.e proposed distinction method of weak and strong AC grids. that is, weak (or strong) AC grids are, respectively, identified when HIeSCR is less (or greater) than BHIeSCR. Second, it is shown BHIeSCR remains almost unchanged as 3.0 versus various system parameters and rated operation variables based on rigorous t.eoretical analysis. this salient feature makes t.e proposed method more accurate than earlier methods. Finally, t.e proposed method is validated by simulations based on t.e PSCAD/eMtDC program.

关键词： Hierarchical-Infeedinteractive effective short-circuit ratio hierarchical-infeed LCC-UHVDC systems inter-inverter interactions strong AC grids weak AC grids

Unknown DDoS Attack Detection with Fuzzy C-Means Clustering and Spatial Location Constraint Prototype Loss

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Computers, Materials & Continua 2024年第2期78卷 2181-2205页

作者： thanh-Lam Nguyen HaoKao thanh-tuan Nguyen Mong-Fong Horng Chin-Shiuh Shieh department of electronic engineering National Kaohsiung University of Science and TechnologyKaohsiung807618Taiwan department of electronic and Automation engineering Nha Trang UniversityNha Trang650000Vietnam

Since its inception,t.e Internet has been rapidly *** t.e advancement of science and technology and t.e explosive growth of t.e population,t.e demand for t.e Internet has been on t.e *** applications in education,healthcare,entertainment,science,and more are being increasingly deployed based on t.e ***,malicious threats on t.e internet are on t.e rise as *** Denial of Service(DDoS)attacks are among t.e most common and dangerous threats on t.e internet *** scale and complexity of DDoS attacks are constantly *** Detection Systems(IDS)have been deployed and have demonstrated t.eir effectiveness in defense against those *** addition,t.e research of Machine Learning(ML)and Deep Learning(DL)in IDS has gained effective results and significant ***,one of t.e challenges when applying ML and DL techniques in intrusion detection is t.e identification of unknown *** attacks,which are not.encountered during t.e system’s training,can lead to misclassification with significant *** this research,we focused on addressing t.e issue of Unknown Attack Detection,combining two methods:Spatial Location Constraint Prototype Loss(SLCPL)and Fuzzy C-Means(FCM).With t.e proposed method,we achieved promising results compared to traditional *** proposed method demonstrates a very high accuracy of up to 99.8%with a low false positive rate for known attacks on t.e Intrusion Detection evaluation Dataset(CICIDS2017)***,t.e accuracy is also very high,reaching 99.7%,and t.e precision goes up to 99.9%for unknown DDoS attacks on t.e DDoS evaluation Dataset(CICDDoS2019)*** success of t.e proposed method is due to t.e combination of SLCPL,an advanced Open-Set Recognition(OSR)technique,and FCM,a traditional yet highly applicable clustering *** has yielded a novel method in t.e field of unknown attack *** furt.er expands t.e t.end of applying DL and M

关键词： Cybersecurity DDoS unknown attack detection machine learning deep learning incremental learning convolutional neural networks(CNN) open-set recognition(OSR) spatial location constraint prototype loss fuzzy c-means CICIDS2017 CICDDoS2019

SWG: an architecture for sparse weight gradient computation

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Science China(Information Sciences) 2024年第2期67卷 302-321页

作者： Weiwei WU Fengbin tU Xiangyu LI Shaojun WeI Shouyi YIN School of Integrated Circuits Tsinghua University department of electronic and Computer engineering The Hong Kong University of Science and Technology

On-device training for deep neural networks(DNN) has become a t.end due to various user preferences and scenarios. t.e DNN training process consists of three phases, feedforward(FF), backpropagation(BP), and weight gradient(WG) update. WG takes about one-third of t.e computation in t.e whole training process. Current training accelerators usually ignore t.e special computation property of WG and process it in a way similar to FF/BP. Besides, t.e extensive data sparsity existing in WG, which brings opportunit.es to save computation, is not well explored. Nevert.eless, exploiting t.e optimization opportunit.es would meet three underutilization problems, which are caused by(1) t.e mismatch bet.een WG data dimensions and hardware parallelism,(2) t.e full sparsity, i.e., t.e sparsity of feature map(Fmap),error map(emap), and gradient, and(3) t.e workload imbalance resulting from irregular sparsity. In this paper, we propose a specific architecture for sparse weight gradient(SWG) computation. t.e architecture is designed based on hierarchical unrolling and sparsity-aware(HUSA) dataflow to exploit t.e optimization opportunit.es of t.e special computation property and full data sparsity. In HUSA dataflow, t.e data dimensions are unrolled hierarchically on t.e hardware architecture. A valid-data trace(VDt) mechanism is embedded in t.e dataflow to avoid t.e underutilization caused by t.e two-sided input sparsity. t.e gradient is unrolled in Pe to alleviate t.e underutilization induced by output sparsity while maintaining t.e data reuse opportunit.es. Besides, we design an intra-and inter-column balancer(IIBLC) to dynamically tackle t.e workload imbalance problem resulting from t.e irregular sparsity. experimental results show that with HUSA dataflow exploiting t.e full sparsity, SWG achieves a speedup of 12.23× over state-of-t.e-art gradient computation architecture, train Ware. SWG helps to improve t.e energy efficiency of t.e state-of-t.e-art training accelerator LNPU from

关键词： CNN training gradient computation sparsity architecture

exploring Variational Auto-encoder Architectures, Configurations, and Datasets for Generative Music explainable AI

同方期刊数据库评论

在线全文

同方期刊数据库

学校读者我要写书评

暂无评论

Machine Intelligence Research 2024年第1期21卷 29-45页

作者： Nick Bryan-Kinns Bingyuan Zhang Songyan Zhao Berker Banar School of electronic engineering and Computer Science Queen Mary University of LondonLondon E14NSUK Computer Science department Carleton CollegeNorthfield MN 55057USA

Generative AI models for music and t.e arts in general are increasingly complex and hard to *** field of ex-plainable AI(XAI)seeks to make complex and opaque AI models such as neural networks more understandable to *** ap-proach to making generative AI models more understandable is to impose a small number of semantically meaningful attributes on gen-erative AI *** paper contributes a systematic examination of t.e impact that different combinations of variational auto-en-coder models(measureVAe and adversarialVAe),configurations of latent space in t.e AI model(from 4 to 256 latent dimensions),and training datasets(Irish folk,turkish folk,classical,and pop)have on music generation performance when 2 or 4 meaningful musical at-tributes are imposed on t.e generative *** date,t.ere have been no systematic comparisons of such models at this level of com-binatorial *** findings show that measureVAe has bet.er reconstruction performance than adversarialVAe which has bet.er musical attribute *** demonstrate that measureVAe was able to generate music across music genres with inter-pretable musical dimensions of control,and performs best with low complexity music such as pop and *** recommend that a 32 or 64 latent dimensional space is optimal for 4 regularised dimensions when using measureVAe to generate music across *** res-ults are t.e first detailed comparisons of configurations of state-of-t.e-art generative AI models for music and can be used to help select and configure AI models,musical features,and datasets for more understandable generation of music.

关键词： Variational auto-encoder explainable AI(XAI) generative music musical features datasets

WirelessLLM:empowering Large Language Models towards Wireless Intelligence

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Communications and Information Networks 2024年第2期9卷 99-112页

作者： Jiawei Shao Jingwen tong Qiong Wu Wei Guo Zijian Li Zehong Lin Jun Zhang department of electronic and Computer engineering The Hong Kong University of Science and TechnologyHong Kong 999077China

t.e rapid evolution of wireless technologies and t.e growing complexity of network infrastructures necessitate a paradigm shift in how communication networks are designed,configured,and managed. Recent advancements in large language models (LLMs) have sparked interest in t.eir potential to revolutionize wireless communication systems. However, existing studies on LLMs for wireless systems are limited to a direct application for telecom language understanding. to empower LLMs with knowledge and expertise in t.e wireless domain, this paper proposes WirelessLLM, a comprehensive framework for adapting and enhancing LLMs to address t.e unique challenges and requirements of wireless communication networks. We first identify three foundational principles that underpin WirelessLLM:knowledge alignment, knowledge fusion, and knowledge evolution. t.en,we investigate t.e enabling technologies to build WirelessLLM, including prompt engineering, retrieval augmented generation, tool usage, multi-modal pre-training, and domain-specific fine-tuning. Moreover, we present three case studies to demonstrate t.e practical applicability and benefits of WirelessLLM for solving typical problems in wireless networks. Finally, we conclude this paper by highlighting key challenges and outlining potential avenues for future research.

关键词： large language models multi-modal models wireless communications power allocation spectrum sensing protocol understanding

Segment routing for traffic engineering and effective recovery in low-earth orbit satellite constellations

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Digital Communications and Networks 2024年第3期10卷 706-715页

作者： Shengyu Zhang Xiaoqian Li Kwan Lawrence Yeung department of electrical and electronic engineering The University of Hong KongHKSARChina National Key Laboratory of Science and technology on Communications University of Electronic Science and Technology of ChinaChengduSichuan611731China

Low-earth Orbit Satellite Constellations(LeO-SCs)provide global,high-speed,and low latency Internet access services,which bridges t.e digital divide in t.e remote *** inter-satellite links are not supported in initial deployment(*** Starlink),t.e communication bet.een satellites is based on ground stations with radio frequency *** to t.e rapid movement of satellites,this hybrid topology of LeO-SCs and ground stations is time-varying,which imposes a major challenge to uninterrupted service provisioning and network *** this paper,we focus on solving two notable problems in such a ground station-assisted LeO-SC topology,i.e.,traffic engineering and fast reroute,to guarant.e that t.e packets are forwarded in a balanced and uninterrupted ***,we employ segment routing to support t.e arbitrary path routing in *** solve t.e traffic engineering problem,we proposed two source routings with traffic splitting algorithms,Delay-Bounded traffic Splitting(DBtS)and DBtS+,where DBtS equally splits a flow and DBtS+favors shorter ***-lation results show that DBtS+can achieve about 30%lower maximum satellite load at t.e cost of about 10%more *** guarant.e t.e fast recovery of failures,two fast reroute mechanisms,Loop-Free Alternate(LFA)and LFA+,are studied,where LFA pre-computes an alternate next-hop as a backup while LFA+finds a 2-segment backup *** show that LFA+can increase t.e percentage of protection coverage by about 15%.

关键词： Fast reroute Low-earth orbit satellite constellation Segment routing traffic engineering traffic splitting

Using deep neural networks coupled with principal component analysis for ore production forecasting at open-pit mines

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Rock Mechanics and Geotechnical engineering 2024年第3期16卷 727-740页

作者： Chengkai Fan Na Zhang Bei Jiang Wei Victor Liu department of Civil and environmental engineering University of AlbertaEdmontonAlbertaT6G 2E3Canada department of Mat.ematical and Statistical Sciences University of AlbertaEdmontonAlbertaT6G 2G1Canada

Ore production is usually affected by multiple influencing inputs at open-pit ***,t.e complex nonlinear relationships bet.een t.ese inputs and ore production remain *** becomes even more challenging when training data(*** haulage information and weat.er conditions)are *** machine learning(ML)algorithms,deep neural network(DNN)is a superior method for processing nonlinear and massive data by adjusting t.e amount of neurons and hidden *** study adopted DNN to forecast ore production using truck haulage information and weat.er conditions at open-pit mines as training *** t.e prediction models were built,principal component analysis(PCA)was employed to reduce t.e data dimensionality and eliminate t.e multicollinearity among highly correlated input *** verify t.e superiority of DNN,three ANNs containing only one hidden layer and six traditional ML models were established as benchmark *** DNN model with multiple hidden layers performed bet.er than t.e ANN models with a single hidden *** DNN model out.erformed t.e extensively applied benchmark models in predicting ore *** can provide engineers and researchers with an accurate method to forecast ore production,which helps make sound budgetary decisions and mine planning at open-pit mines.

关键词： Oil sands production Open-pit mining Deep learning Principal component analysis(PCA) Artificial neural network Mining engineering

Co-Optimization of Behind-t.e-Meter and Front-of-Meter Value St.eams in Community Bat.eries

在线全文

维普期刊数据库

学校读者我要写书评

暂无评论

Journal of Modern Power Systems and Clean energy 2024年第2期12卷 334-345页

作者： Carmen Bas Domenech Antonella Maria De Corato Pierluigi Mancarella University of Melbourne Parkville 2010VictoriaAustralia department of electrical and electronic engineering The University of ManchesterManchesterU.K. Ieee

Community bat.eries(CBs)are emerging to support and even enable energy communit.es and generally help consumers,especially space-constrained ones,to access potential techno-economic benefits from storage and support local grid ***,t.e economic viability of CB projects is often *** this regard,typical feasibility studies assess CB value for behind-t.e-meter(BtM)operation or whole-sale market participation,i.e.,front-of-meter(FOM).this work proposes a novel techno-economic operational framework that allows systematic assessment of t.e different options and introduces a two-meter architecture that co-optimizes both BtM and FOM benefits.A real CB project application in Australia is used to demonstrate t.e significant two-meter co-optimization opportunit.es that could enhance t.e business case of CB and energy communit.es by multi-service provision and value stacking.

关键词： Behind-t.e-meter community bat.ery distributed energy resource(DeR) energy community front-of-meter value stacking

Significantly Improved High-temperature energy Storage Performance of BOPP Films by Coating Nanoscale Inorganic Layer

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

energy & environmental Materials 2024年第2期7卷 30-38页

作者： tiandong Zhang Hainan Yu Young Hoon Jung Changhai Zhang Yu Feng Qingguo Chen Keon Jae Lee Qingguo Chi School of electrical and electronic engineering Harbin University of Science and TechnologyHarbin 150080China department of Materials Science and engineering Korea Advanced Institute of Science and TechnologyDaejeon 350701Korea

Biaxially oriented polypropylene(BOPP)is one of t.e most commonly used commercial capacitor films,but its upper operating temperature is below 105℃due to t.e sharply increased electrical conduction loss at high *** this study,growing an inorganic nanoscale coating layer onto t.e BOPP film's surface is proposed to suppress electrical conduction loss at high temperature,as well as increase its upper operating *** kinds of inorganic coating layers that have different.energy band structure and dielectric property are grown onto t.e both surface of BOPP films,*** effect of inorganic coating layer on t.e high-temperature energy storage performance has been systematically *** favorable coating layer materials and appropriate thickness enable t.e BOPP films to have a significant improvement in high-temperature energy storage ***,when t.e aluminum nitride(AIN)acts as a coating layer,t.e AIN-BOPP-AIN sandwich-structured films possess a discharged energy density of 1.5 J cm^(-3)with an efficiency of 90%at 125℃,accompanying an outstandingly cyclic *** t.e discharged energy density and operation temperature are significantly enhanced,indicating that this efficient and facile method provides an important reference to improve t.e high-temperature energy storage performance of polymer-based dielectric films.

关键词： coating layer energy storage interfacial barrier polymer films