检索结果-内蒙古大学图书馆

ieee International conference on Fuzzy Systems (FUZZ-ieee) / ieee World Congress on Computational Intelligence (ieee WCCI) / International Joint conference on Neural networks (IJCNN) / ieee Congress on Evolutionary Computation (ieee CEC)

作者： Cornell, Filip Karlgren, Jussi Animesh Girdzijauskas, Sarunas KTH Royal Inst Technol Sch Elect Engn & Comp Sci Stockholm Sweden Gavagai Stockholm Sweden Indian Inst Technol Kharagpur Ctr Excellence Artificial Intelligence Kharagpur W Bengal India

ISBN: (数字)9781728186719

ISBN: (纸本)9781728186719

In this paper, we propose a novel way of representing graphs for processing in Graph Neural networks. We reduce the dimensionality of the input data by using Random Indexing, a Vector Symbolic Architectural framework;we implement a new trainable neural layer, also inspired by Vector Symbolic architectures;we leverage the sparseness of the incoming data in a Sparse Neural network framework. Our experiments on a number of publicly available datasets and standard benchmarks demonstrate that we can reduce the number of parameters by up to two orders of magnitude. We show how this parsimonious approach not only delivers competitive results but even improves performance for node classification and link prediction. We find that this holds in particular for cases where the graph lacks node features.

关键词： vector symbolic architectures graph neural networks random indexing

来源：评论

学校读者我要写书评

暂无评论

Innovative Communication Protocols and network architectures in 5G Technologies 3

Innovative Communication Protocols and Network Architectures...

引用

3rd International conference for Advancement in Technology, ICONAT 2024

作者： Senthilkumar, C. Vijay, J. Sankar, R. Akula, Akhila Supriya, S. Purushothaman, V. Saveetha School of Engineering Department of Electronics and Communiation Engineering Tamil Nadu Chennai India Aarupadai Veedu Institute of Technology Vinayaka Mission's Research Foundation Department of Electronics and Communication Engineering Salem India Chennai Institute of Technology Kundrathur Department of EEE Chennai India MLR Institute of Technology Department of ECE Hyderabad India Vel Tech Rangarajan Dr. Sagunthala R&D Institute of Science and Technology Department of Computer Science and Engineering India Chennai Institute of Technology Department of Electronics and Communication Engineering Tamil Nadu Chennai India

ISBN: (纸本)9798350354171

A new era of connectedness marked by unheard-of speed, minimal latency, and increased capacity to serve an abundance of devices and apps is brought about by the introduction of 5G technology. This paper delves into the innovative communication protocols and network architectures that underpin 5G technology, which collectively aim to meet the diverse and demanding requirements of modern communication networks. We investigate the architecture and application of cutting-edge protocols, including improved Mobile Broadband (eMBB), Ultra-Reliable Low-Latency Communications (URLLC), and New Radio (NR). Furthermore, this research delves into innovative network designs such as Massive MIMO, Edge Computing, and network Slicing, emphasizing their functions in streamlining network operations, allocating resources, and augmenting user satisfaction. The integration of these cutting-edge protocols and architectures within the 5G framework promises to achieve higher efficiency, flexibility, and scalability. However, the deployment of 5 G networks is not without challenges. This paper addresses key issues such as security vulnerabilities, interoperability among heterogeneous devices and networks, and the substantial investments required for infrastructure development. Additionally, we discuss the regulatory and standardization hurdles that must be overcome to realize the full potential of 5G technology. Using a thorough study, we highlight how 5G technology is revolutionizing a number of industries, including telemedicine, the Internet of Things (IoT), driverless cars, and smart cities. The last section of the paper paints a picture of how global connectivity will develop in the future, with 5G acting as a key technology that will open the door to a more intelligent, connected, and productive society. © 2024 ieee.

关键词： Internet of things

来源：评论

学校读者我要写书评

暂无评论

Optimal Planning of Mobile Battery Energy Storage in Distribution network with Soft open Point

Optimal Planning of Mobile Battery Energy Storage in Distrib...

引用

58th ieee/IAS Industrial and Commercial Power Systems Technical conference Asia (ieee I and CPS Asia)

作者： Liu, Tianqi Chen, Jian Zhang, Yicheng Shandong Univ Key Lab Power Syst Intelligent Dispatch & Control Minist Educ Jinan Peoples R China Agency Sci Technol & Res ASTAR Inst Infocomm Res 12R Singapore Singapore

ISBN: (数字)9781665450669

ISBN: (纸本)9781665450669

To address the need for high-quality power, the distribution network (DN) is gradually incorporating battery energy storage (BES) and flexible interconnection equipment, such as Soft open Point (SOP). This paper proposes a mobile battery energy storage (MBES) planning strategy considering multiple scenarios in DN with SOP. Firstly, the fundamental concepts of commonly used equipment like SOP and energy storage are investigated and mathematical models are constructed. Secondly, taking the minimum annual comprehensive investment and network loss cost of DN as the planning goal, a MBES planning model for promoting the economics of the DN under multiple scenarios is established. The nonlinear programming model is transformed into a second order cone model by linearization and convex relaxation, which can achieve an accurate and rapid solution. Finally, the improved ieee-33 node example is used for simulation, and the MBES and BES planning approaches are compared. Results show that combining MBES with SOP control can not only enhance system efficiency but also reduce voltage fluctuation, demonstrating the effectiveness of the suggested technique.

关键词： distribution network soft open point mobile battery energy storage second order cone programming

来源：评论

学校读者我要写书评

暂无评论

HEROv2: Full-Stack open-Source Research Platform for Heterogeneous Computing

引用

ieee TRANSACTIONS ON PARALLEL AND DISTRIBUTED SYSTEMS 2022年第12期33卷 4368-4382页

作者： Kurth, Andreas Forsberg, Bjoern Benini, Luca Integrated Syst Lab CH-8092 Zurich Switzerland Swiss Fed Inst Technol D ITET Dept Informat Technol & Elektrotech CH-8092 Zurich Switzerland Univ Bologna DEIS I-40126 Bologna BO Italy

Heterogeneous computers integrate general-purpose host processors with domain-specific accelerators to combine versatility with efficiency and high performance. To realize the full potential of heterogeneous computers, however, many hardware and software design challenges have to be overcome. While architectural and system simulators can be used to analyze heterogeneous computers, they are faced with unavoidable compromises between simulation speed and performance modeling accuracy. In this work we present HEROv2, an FPGA-based research platform that enables accurate and fast exploration of heterogeneous computers consisting of accelerators based on clusters of 32-bit RISC-V cores and an application-class 64-bit ARMv8 or RV64 host processor. HEROv2 allows to seamlessly share data between 64-bit host s and 32-bit accelerators and comes with a fully open-source on-chip network, a unified heterogeneous programming interface, and a mixed-data-model, mixed-ISA heterogeneous compiler based on LLVM. We evaluate HEROv2 in four case studies from the application level over toolchain and system architecture down to accelerator microarchitecture. We demonstrate how HEROv2 enables effective research and development on the full stack of heterogeneous computing. For instance, the compiler can tile loops and infer data transfers to and from the accelerators, which leads to a speedup of up to 4.4 x compared to the original program and in most cases is only 15% slower than a handwritten implementation, which requires 2.6 x more code.

关键词： Hardware Program processors Codes Computers System-on-chip programming Engines Heterogeneous (Hybrid) systems parallel processors shared memory hardware software interfaces system architectures

来源：评论

学校读者我要写书评

暂无评论

Emulation Learning for Neuromimetic Systems 62

Emulation Learning for Neuromimetic Systems

引用

62nd ieee conference on Decision and Control (CDC)

作者： Sun, Zexin Baillieul, John Boston Univ Div Syst Engn Boston MA 02215 USA Boston Univ Dept Mech Engn Boston MA 02215 USA Boston Univ Dept Elect & Comp Engn Boston MA 02215 USA

ISBN: (纸本)9798350301243

Building on our recent research on neural heuristic quantization systems, results on learning quantized motions and resilience to channel dropouts are reported. We propose a general emulation problem consistent with the neuromimetic paradigm. This optimal quantization problem can be solved by model predictive control (MPC), but because the optimization step involves integer programming, the approach suffers from combinatorial complexity when the number of input channels becomes large. Even if we collect data to train a neural network, collection of training data and the training itself are still time-consuming. Therefore, we propose a general Deep Q network (DQN) algorithm that not only learns trajectories but also exhibits the advantages of resilience to channel dropout. Furthermore, to transfer the model to other emulation problems, a mapping-based transfer learning approach can be used directly on the current model to obtain the optimal trajectory steps for the new emulation problems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Accelerating Model Solving for Integrated Optimization of Timetabling and Vehicle Scheduling based on Graph Convolutional network 26

Accelerating Model Solving for Integrated Optimization of Ti...

引用

ieee 26th International conference on Intelligent Transportation Systems (ITSC)

作者： Liu, YeCheng Chen, Xu Xu, YongXin Xiang, Dong Mo, LinJian Space 569 Xixi Rd Hangzhou Peoples R China

ISBN: (纸本)9798350399462

We study the integrated optimization of timetabling and vehicle scheduling for typical two-way bus lines in public bus transportation. A novel Mixed Integer Linear programming (MILP) model is built for the integrated problem, where some real-life and complex operational constraints, e.g., the upper limits of headways in different time periods, lunch breaks, and max working time, are considered. Since the integrated problem itself is NP-Hard and is difficult to be solved efficiently by off-the-shelf open source solvers, we especially propose a novel approach based on Graph Convolutional network(GCN) to explore the correlations between problem structures and the resulting solution values. In this methodology, an integrated problem instance is represented as a trip chain graph, and a supervised GCN is trained to predict the solution value of partial binary variables which are hard to determine during problem solving. The values of predicted variables will be fixed and the remaining subMILP with smaller search space can be solved faster than the original MILP. To our knowledge, this is the first time machine learning techniques have been used to accelerate the joint solving of timetabling and vehicle scheduling in literature. Our computational evaluations of 100 real problem instances show that the GCN-based methodology outperforms existing solvers, reducing average wall-clock time by 20% while achieving better solution quality.

关键词： Learning systems

来源：评论

学校读者我要写书评

暂无评论

Evolutionary Multitasking CNN Architecture Search for Hyperspectral Image Classification

Evolutionary Multitasking CNN Architecture Search for Hypers...

引用

作者： Liu, Yiting Li, Hao Gong, Maoguo Liu, Jieyi Wu, Yue Zhang, Mingyang Shi, Jiao Xidian Univ Sch Elect Engn Xian Peoples R China Northwestern Polytech Univ Xian Peoples R China

ISBN: (数字)9781728186719

ISBN: (纸本)9781728186719

In recent years, convolutional neural networks (CNNs) have shown excellent effectiveness on hyperspectral image classification (HSI) tasks. However, it is a challenge to design a suitable CNN architecture to obtain great performance according to different tasks. Different from the traditional manual design, in this paper, an evolutionary multitasking CNN architecture search framework for HSI classification is proposed to search the optimal architectures and accomplish classification of different tasks simultaneously. Through encoding the CNN architectures, the proposed algorithm is able to achieve global search in the same search space and select well-adapted individuals for evolution. In the evolutionary multitasking environment, information can be transferred between and within tasks, which can accelerate the convergence and explore good architectures through beneficial transfer. In the experiments, the effectiveness of the proposed method is demonstrated by the comparison with different methods on two common data sets.

关键词： Hyperspectral images classification convolutional neural network evolutionary multitasking

来源：评论

学校读者我要写书评

暂无评论

FGA-NAS: Fast Resource-Constrained Architecture Search by Greedy-ADMM Algorithm

FGA-NAS: Fast Resource-Constrained Architecture Search by Gr...

引用

作者： Chen, Yifei Zhang, Junge Li, Qiaozhe Chen, Hao Huang, Kaiqi Univ Chinese Acad Sci Sch Artificial Intelligence Beijing Peoples R China Chinese Acad Sci Inst Automat CRISE Beijing Peoples R China

ISBN: (数字)9781728186719

ISBN: (纸本)9781728186719

Differentiable architecture search has demonstrated promising results in automatically designing neural network architectures with desired properties, such as high accuracy and low FLOPs. However, it suffers from a cumbersome training process, and the injection of constraints in the search phase often relies on some hand-crafted heuristic regularizers, the design of which typically requires tremendous human effort. In this paper, to address these critical challenges, we present FGA-NAS, an efficient method for resource-constrained architecture search. First, to reduce the computational cost and improve search flexibility, we propose a novel condensed search space that merges multiple parallel-placed candidates into a single one. Second, to enable the gradient-based optimization for neural architecture search (NAS) under multiple combinatorial constraints, we decompose the constrained NAS into a few simple sub-problems without introducing any heuristics by using the ADMM algorithm [1]. Then, the constrained NAS can be resolved by alternately solving the simple sub-problems. Experimental results on ImageNet show that our method can discover efficient and accurate neural network architectures that achieve the state-of-the-art by only using 0.2 GPU days.

关键词： Deep Learning Convolutional Neural network Neural Architecture Search

来源：评论

学校读者我要写书评

暂无评论

NeXtCrowd: Lightweight And Efficient network Design for Dense Crowd Counting 25

NeXtCrowd: Lightweight And Efficient Network Design for Dens...

引用

25th ieee International conferences on High Performance Computing and Communications, 9th International conference on Data Science and Systems, 21st ieee International conference on Smart City and 9th ieee International conference on Dependability in Sensor, Cloud and Big Data Systems and Applications, HPCC/DSS/SmartCity/DependSys 2023

作者： Hu, Jun Han, Hui College of Computer Science and Electronic Engineering Hunan University China Key Laboratory of Natural Resources Monitoring and Supervision in Southern Hilly Region Ministry of Natural Resources China

ISBN: (纸本)9798350330014

Dense crowd counting remains a challenging task due to complex backgrounds, large-scale scenes, diversity of human features, and computational performance requirements. Unlike past solutions with additional designs or complex architectures, we argue that the counting task only needs to give an approximate number or density distribution of people in the scene, without exact location and classification, suggesting that a relatively lightweight architecture is sufficient for this task. Motivated by this line of thought, we explore whether a simple convolutional architecture can maintain counting performance in dense crowd counting. We propose an efficient CNN framework for crowd counting, namely NeXtCrowd. Specifically, we eliminate some unnecessary architectures to reduce the complexity and number of parameters of the backbone architecture, thus improving its computational efficiency and generalization capability. We then improve the performance and stability of the network by choosing the corresponding loss function and optimization method based on the task characteristics. Last we adopt a regression layer with Global Response Normalization to generate the density map, which can make better use of the information of each channel in the deep image features, and aggregate a more accurate density map of dense crowd. Without using additional architectures, NeXtCrowd can still achieve state-of-the-art performance on several dense crowd datasets. On the UCF-QNRF and ShanghaiTech Part A datasets, our model reduces the root mean squared error by 16.87 (131.4) and 7.9 (86.29), respectively' over the baseline methods. © 2023 ieee.

关键词： Computer vision

来源：评论

学校读者我要写书评

暂无评论

Power System Monitoring Through Low-Voltage Distribution network Using freePMU

引用

ieee TRANSACTIONS ON INDUSTRY APPLICATIONS 2022年第3期58卷 3153-3163页

作者： Seger, Pedro V. H. Grando, Flavio L. Lazzaretti, Andre E. Moreto, Miguel Denardin, Gustavo W. Pastro, Cristian R. Univ Fed Santa Catarina BR-88040900 Florianopolis SC Brazil Univ Tecnol Fed Parana BR-80230901 Curitiba Parana Brazil

This article presents the latest developments of a phasor measurement unit (PMU), hereby named freePMU, aimed at a large-scale deployment in low-voltage distribution networks. The defining characteristics of this device are its low cost, general-purpose hardware architecture, and software now made open source. This PMU has already been shown to reliably measure voltage, angle, and frequency events in a real distribution network while complying with the ieee class P PMU standards. In this work, several new results are shown, such as harmonic estimation, detection of different consumer load patterns, and load variations on a distribution feeder. Moreover, its hardware and software architectures are now made public, contributing to the project goal of a cheap and mass-deployable PMU. All of those results attest to the original contributions of this work, compared to other PMUs focused on distribution systems.

关键词： Phasor measurement units Estimation Monitoring Harmonic analysis Global Positioning System Synchronization Prototypes Distribution network monitoring harmonic synchrophasor phasor estimation phasor measurement PMU

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：