检索结果-内蒙古大学图书馆

High Precision and Fast distributed Temperature Data Demodulation Algorithm of Optical Frequency Domain Reflectometer Based on LSTM-CNN

引用

IEEE JOURNAL OF QUANTUM ELECTRONICS 2024年第6期60卷 1页

作者： Lei, Huang Min, Liu Cui, Yingqi Zhu, Zhaohao Ping, Shum Chongqing Univ Sch Microelect & Commun Engn Chongqing 400044 Peoples R China Southern Univ Sci & Technol Dept Elect & Elect Engn Shenzhen 518055 Peoples R China

A demodulation algorithm based on the LSTM-CNN is proposed to simultaneously achieve the demodulation of temperature data from distributed optical frequency domain reflectometry (OFDR). As for the local measurement range along the distributed fiber, the LSTM-CNN can achieve an average mean absolutely error (MAE) of only 0.0393 and the average demodulation time is only 0.1507 seconds. The comparison with the cross-correlation algorithm, Multi-Layer Perceptron (MLP), Extreme Learning Machine (ELM), Long Short-Term Memory (LSTM), and Convolutional neural network (CNN) demonstrates that the MAE is reduced by 85.98%, 77.23%, 88.25%, 80.95%, and 91.82%, and the average time is faster 38.19 times, 8.71 times, 3.28 times, 1.37 times, and 2.45 times, respectively. As for the full measurement range of the distributed fiber, the temperature distribution curve demodulated by LSTM-CNN is found to be consistent with the actual temperature distribution curve and the average demodulation time is 0.371 seconds, providing a new method for the temperature data demodulation in the distributed OFDR sensing system.

关键词： Temperature measurement Temperature distribution Demodulation Temperature sensors Training Long short term memory Sensors Optical fiber networks Convolutional neural networks Feature extraction Optical frequency domain reflectometer machine learning LSTM-CNN data processing

来源：评论

学校读者我要写书评

暂无评论

A collaborative cloud-edge computing framework in distributed neural network

引用

EURASIP JOURNAL ON WIRELESS COMMUNICATIONS AND networkING 2020年第1期2020卷 1-17页

作者： Xu, Shihao Zhang, Zhenjiang Kadoch, Michel Cheriet, Mohamed Beijing Jiaotong Univ Beijing Municipal Commiss Educ Key Lab Commun & Informat Syst Sch Elect & Informat Engn Beijing Peoples R China Beijing Jiaotong Univ Sch Software Engn Beijing 100044 Peoples R China Ecole Technol Super ETS Montreal PQ H3C 1K3 Canada

The emergence of edge computing provides a new solution to big data processing in the Internet of Things (IoT) environment. By combining edge computing with deep neural network, it can make better use of the advantages of multi-layer architecture of the network. However, the current task offloading and scheduling frameworks for edge computing are not well applicable to neural network training tasks. In this paper, we propose a task model offloading algorithm by considering how to optimally deploy neural network model into the edge nodes. An adaptive task scheduling algorithm is also designed to adaptively optimize the task assignment by using the improved ant colony algorithm. Based on them, a collaborative cloud-edge computing framework is proposed, which can be used in the distributed neural network. Moreover, this framework sets up some mechanisms so that the cloud can collaborate with edge computing in the work. The simulation results show that the framework can reduce time delay and energy consumption, and improve task accuracy.

关键词： Edge computing distributed neural network Resource allocation Task offloading

来源：评论

学校读者我要写书评

暂无评论

Split Edge-Cloud neural networks for Better Adversarial Robustness

引用

IEEE ACCESS 2024年 12卷 158854-158865页

作者： Douch, Salmane Abid, Mohamed Riduan Zine-Dine, Khalid Bouzidi, Driss Benhaddou, Driss Mohammed V Univ Rabat Natl Sch Comp Sci & Syst Anal ENSIAS Rabat 30050 Morocco Columbus State Univ TSYS Sch Comp Sci Columbus GA 31907 USA Mohammed V Univ Rabat Fac Sci FSR Rabat 30050 Morocco Alfaisal Univ Coll Engn Riyadh 11533 Saudi Arabia

Cloud computing is a critical component in the success of 5G and 6G networks, particularly given the computation-intensive nature of emerging applications. Despite all it advantages, cloud computing faces limitations in meeting the strict latency and bandwidth requirements of applications such as eHealth and automotive systems. To overcome these limitations, edge computing has emerged as a novel paradigm that bring computation closer to the user. Moreover, intelligent tasks such as deep learning ones demand more memory and processing power than edge devices can handle. To address these challenges, methods like quantization, pruning, and distributed inference have been proposed. Similarly, this paper study a promising approach for running deep learning models at the edge: split neural networks (SNN). SNNs feature a neural network architecture with multiple early exit points, allowing the model to make confident decisions at earlier layers without processing the entire network. This not only reduces memory and computational demands but it also makes SNNs well-suited for edge computing applications. As the use of SNNs expands, ensuring their safety-particularly their robustness to perturbations-becomes crucial for deployment in safety-critical scenarios. This paper presents the first in-depth study on the robustness of split Edge Cloud neural networks. We review state-of-the-art robustness certification techniques and evaluate SNN robustness using the auto_LiRPA and Auto Attack libraries, comparing them to standard neural networks. Our results demonstrate that SNNs reduce average inference time by 75'% and certify 4 to 10 times more images as robust, while improving overall robustness accuracy by 1% to 10%.

关键词： Robustness Edge computing Perturbation methods Computational modeling Cloud computing Certification Biological neural networks Quantization (signal) Image edge detection Deep learning Adversarial attacks cloud computing edge computing edge intelligence robustness certification split neural networks

来源：评论

学校读者我要写书评

暂无评论

PAICORE: A 1.9-Million-Neuron 5.181-TSOPS/W Digital Neuromorphic Processor With Unified SNN-ANN and On-Chip Learning Paradigm

引用

IEEE JOURNAL OF SOLID-STATE CIRCUITS 2025年第2期60卷 651-671页

作者： Zhong, Yi Kuang, Yisong Liu, Kefei Wang, Zilin Feng, Shuo Chen, Guang Yang, Youming Cui, Xiuping Wang, Qiankun Cao, Jian Jia, Song Liang, Yun Sun, Guangyu Cui, Xiaoxin Huang, Ru Wang, Yuan Peking Univ MPW Ctr Sch Integrated Circuits Key Lab Microelect Devices & Circuits Beijing 100871 Peoples R China Beijing Adv Innovat Ctr Integrated Circuits Beijing 100871 Peoples R China

The neuromorphic approach of fulfilling brain-like edge intelligence is confronted with three paramount challenges: 1) ever-increasing application demands versus insufficient on-chip resources;2) diverse data sources and models versus heterogeneous computing paradigms;and 3) adaptation to real scenarios versus absence of on-chip learning. To bridge those gaps in the current research community, this article presents PAICORE, a 537.98-mm(2) digital neuromorphic processor with a unified computing and learning paradigm. PAICORE is a scalable 1024-core design, integrating over 1.919 million neurons and 4.773 billion synapses on a single chip. PAICORE is implemented in globally asynchronous locally synchronous (GALS) style with a five-level fat up-down quadtree as dedicated network-on-chip (NoC) infrastructure and tiled 2-D chip array as inter-chip architecture. The distributed processing cores fuse the hybrid spiking neural network (SNN), artificial neural network (ANN), and on-chip spike-timing-dependent plasticity (STDP) learning models into a unified description by modular-level reconfiguration. PAICORE achieves the peak performance of 20.74 TSOPS, 41.49 TOPS, and 115.25 GSOPS, with the best energy efficiency of 5.181 TSOPS/W, 10.372 TOPS/W, and 1.222 TSOPS/W for SNN, ANN, and on-chip learning paradigms, respectively. Benefiting from the PAIFLOW software framework, PAICORE hardware platform can be equivalently simulated and efficiently programmed in a large-scale deployment, in accordance with specific optimizing demands for its target multi-paradigm tasks.

关键词： Neurons Computational modeling Artificial neural networks Biological system modeling Data models Adaptation models Task analysis Artificial neural network (ANN) edge application multi-chip scalability network-on-chip (NoC) neuromorphic computing on-chip learning spike-timing-dependent plasticity (STDP) spiking neural network (SNN) unified paradigm

来源：评论

学校读者我要写书评

暂无评论

distributed Manipulability optimization in a Finite Time neural network for Redundant Manipulators 11

Distributed Manipulability optimization in a Finite Time Neu...

引用

11th International Conference on Intelligent Control and Information processing, ICICIP 2021

作者： Kong, Ying Wu, Jiajia Chen, Shiyong Zhou, Junwen Zhejiang University of Science and Technology Department of Information and Electronic Engineering China

ISBN: (纸本)9781665425155

A distributed manipulability optimization (DMO) scheme based on a finite time neural network is proposed in this paper to solve the cooperative motion planning of redundant manipulators. In this proposed kinematic scheme, the end-effectors of the manipulators can complete the specific task in a cooperative manner under peer-to-peer communication and the optimal kinematic time of redundant manipulators has achieved. The DMO scheme is formulated into a quadratic program and is solved by Lagrange multiplier theorem. The stability and finiteness of the proposed DMO scheme have been proved in theory. Simulation results on three redundant manipulators show the validity and accuracy of this new DMO scheme. © 2021 IEEE.

关键词： Redundant manipulators

来源：评论

学校读者我要写书评

暂无评论

Identifying the neural Bases of Math Competence Based on Structural and Functional Properties of the Human Brain

引用

JOURNAL OF COGNITIVE NEUROSCIENCE 2023年第8期35卷 1212-1228页

作者： Ren, Xueying Libertus, Melissa E. Univ Pittsburgh Pittsburgh PA 15260 USA

Human populations show large individual differences in math performance and math learning abilities. Early math skill acquisition is critical for providing the foundation for higher quantitative skill acquisition and succeeding in modern society. However, the neural bases underlying individual differences in math competence remain unclear. Modern neuroimaging techniques allow us to not only identify distinct local cortical regions but also investigate large-scale neural networks underlying math competence both structurally and functionally. To gain insights into the neural bases of math competence, this review provides an overview of the structural and functional neural markers for math competence in both typical and atypical populations of children and adults. Although including discussion of arithmetic skills in children, this review primarily focuses on the neural markers associated with complex math skills. Basic number comprehension and number comparison skills are outside the scope of this review. By synthesizing current research findings, we conclude that neural markers related to math competence are not confined to one particular region;rather, they are characterized by a distributed and interconnected network of regions across the brain, primarily focused on frontal and parietal cortices. Given that human brain is a complex network organized to minimize the cost of information processing, an efficient brain is capable of integrating information from different regions and coordinating the activity of various brain regions in a manner that maximizes the overall efficiency of the network to achieve the goal. We end by proposing that frontoparietal network efficiency is critical for math competence, which enables the recruitment of task-relevant neural resources and the engagement of distributed neural circuits in a goal-oriented manner. Thus, it will be important for future studies to not only examine brain activation patterns of discrete regions but also exam

关键词： Brain

来源：评论

学校读者我要写书评

暂无评论

WLAN Channel Status Duration Prediction for Audio and Video Services Using Probabilistic neural networks

引用

IEEE ACCESS 2024年 12卷 28201-28211页

作者： Hou, Yafei Denno, Satoshi Okayama Univ Fac Environm Life Nat Sci & Technol Okayama 7008530 Japan

Due to massive increase in wireless access from smartphones, IoT devices, WLAN is aiming to improve its spectrum efficiency (SE) using many technologies. Some interesting techniques for WLAN systems are flexible allocation of frequency resource and cognitive radio (CR) techniques which expect to find more useful spectrum resource by modeling and then predicting of channel status using the captured statistics information of the used spectrum. This paper investigates the prediction accuracy of busy/idle duration of two major wireless services: audio service and video service using neural network based predictor. We first study the statistics distribution of their time-series busy/idle (B/I) duration, and then analyze the predictability of the busy/idle duration based on the predictability theory. Then, we propose a data categorization (DC) method which categorizes the duration of recent B/I duration according the their ranges to make the duration of next data be distributed into several streams. From the predictability analysis of each stream and the prediction performance using the probabilistic neural network (PNN), it can be confirmed that the proposed DC can improve the prediction accuracy of time-series data in partial streams.

关键词： Wireless LAN Wireless communication Media streaming Wireless sensor networks Resource management Probability distribution Channel allocation Audio-visual systems Data processing Predictive models neural networks Channel status duration prediction WLAN audio/video traffic data predictability analysis probabilistic neural network (PNN)

来源：评论

学校读者我要写书评

暂无评论

Fault Detection and Fault Location in a Grid-Connected Microgrid Using Optimized Deep Learning neural network

引用

OPTIMAL CONTROL APPLICATIONS & METHODS 2025年第3期46卷 896-911页

作者： Karthick, R. Saravanan, R. Arulkumar, P. SNS Coll Technol Dept Elect & Elect Engn Coimbatore Tamilnadu India Balaji Inst Technol & Sci Dept Elect & Elect Engn Warangal Telangana India VSB Engn Coll Dept Elect & Elect Engn Karur Tamil Nadu India

The significant prevalence of distributed energy resources in microgrids due to their unique characteristics and activities creates protection issues. This paper introduces fault detection and its location in an MG. The aim of the investigation is to enhance the system's efficiency and dependability, and fault detection and prediction are employed. Data are gathered from the operational Microgrid (MG) infrastructure dataset, and the data are fed to the Newton Time-Extracting Wavelet Transform (NTEWT). It cleanses the data and enhances the input data's quality. After the preprocessed data are fed to the feature extraction by using Distorted Gaussian Matched Filtering (DGMF), it extracts the five statistical features from the voltage signal. The optimal features are fed to the Deep Attention Dilated Residual Convolutional neural network (DADRCNN), which predicts the faults in the microgrid. The Dung Beetle Optimization (DBO) is utilized to optimize the weight parameter of DADRCNN. The MATLAB platform is utilized to implement the proposed method. The proposed method is contrasted with various existing approaches like Graph Convolutional network and Graph Fourier Transform (GCN-GFT), Feed-Forward neural network and Back-Propagation Algorithm (FFNN-BPA), and Discrete Wavelet Transform and Radial Basis Function neural network (DWT-RBFN). The existing method shows a fault detection rate of 91.12, 88.80, and 84.239, and the proposed technique illustrates a fault detection rate of 99.9640, which is higher than other existing approaches. The existing method shows a false alarm rate of 0.758, 1.028, and 1.564, and the proposed technique shows a false alarm rate of 0.2519, which is lower than other existing methods. The proposed approach exhibits a reduced false alarm rate and a greater fault detection rate, it is concluded.

关键词： data processing data set deep learning fault detection microgrid microgrid infrastructure

来源：评论

学校读者我要写书评

暂无评论

Data processing for object recognition from satellite images using convolutional neural networks: a case study in solar panel recognition with FORMOSAT-2

引用

JOURNAL OF APPLIED REMOTE SENSING 2024年第3期18卷

作者： Chen, Bo-Wei Hsu, Hwai-Jung Chang, Yu-Yun Liu, Cynthia S. J. Huang, Winfred Feng Chia Univ Dept Informat Engn & Comp Sci Taichung Taiwan Feng Chia Univ Res Ctr Artificial Intelligence Taichung Taiwan Taiwan Space Agcy Hsinchu Taiwan

Investigating ground objects widely distributed in geography and large in scale is one of the primary missions for satellite sensors. On the other hand, recognizing objects from images is one of the classic tasks for convolutional neural networks (CNNs), currently the most popular computer vision technique. Data processing, such as data augmentation, channel selection, and image fusion, can be essential when applying CNNs to satellite images. With a case study of recognizing solar panels from satellite images using CNN, the related data processing issues are discussed, and an approach to embed channel fusion methods into CNN is established. As a result, the following findings are concluded from our case study: (1) not all channels in satellite images contribute to specific object recognition, and thus channel selection is necessary in applying CNN on satellite images;(2) fine-tuning the fusion method embedded in CNN improves the model stability;and (3) transfer learning is outperformed by CNN models trained with augmented data for object recognition from satellite images.

关键词： convolutional neural network object recognition satellite images transfer learning

来源：评论

学校读者我要写书评

暂无评论

Modeling and distributed adaptive fault-tolerant vibration control for bridge beam with single-parameter adaptive neural network

引用

INTERNATIONAL JOURNAL OF ADAPTIVE CONTROL AND SIGNAL processing 2020年第12期34卷 1831-1846页

作者： Gao, Shiqi Yang, Hongjun Liu, Jinkun Beihang Univ Sch Automat Sci & Elect Engn Xueyuan Rd 37 Beijing 100191 Peoples R China Chinese Acad Sci Inst Automat Beijing Peoples R China

Modeling and vibration control of a bridge beam system are considered in this article. The beam bridge with both ends fixed can be regarded as an Euler-Bernoulli beam, which is a typical distributed parameter system. First, the partial differential equations (PDE) model of the bridge was established according to the Hamilton principle. Then, a reasonable distributed control law was designed on the PDE model to eliminate the elastic deformation and suppress the vibration of the bridge. At the same time, uncertainties related to system status were considered during the design of the closed-loop system. In addition, the possible actuator and sensor faults in the control system were analyzed. Single-parameter adaptive neural networks were used to estimate the effects of coupling terms for uncertainties and faults. The parameter estimation adaptive law was designed to replace the adjustment of neural network weights, which simplifies the algorithm and facilitates practical engineering applications. Finally, the feasibility of the control system was verified by simulation.

关键词： actuator fault distributed control distributed parameter system sensor fault single-parameter neural network vibration control

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：