检索结果-内蒙古大学图书馆

OCEANS

作者： Harel Mendelman Tomer Masssas Yair Moshe Signal and Image Processing Laboratory (SIPL) Andrew and Erna Viterbi Faculty of Electrical & Computer Engineering Technion – Israel Institute of Technology Haifa Israel

Automatic monitoring of beach video streams is important for improving safety, environmental monitoring, research, and education related to beach activities. This paper introduces a novel approach for monitoring water bodies by analyzing beach video streams. In contrast to earlier works, in the proposed approach we analyze not only the behavior of water bodies, or of humans on beach videos, but also the interactions between them. By integrating human activity and water behavior analysis, the approach provides new insights that are unattainable by analyzing each separately. To accomplish this objective, deep neural networks are utilized to analyze video streams from existing beach webcams. The analysis includes the detection and tracking of people and waves, as well as higher level analysis. We use the task of characterizing surfing conditions as a case study and demonstrate our ability to estimate the values of key parameters that can help determine the quality of a surf spot at a given time.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Reinforcement Learning-Enhanced Active Disturbance Rejection Control for Mobile Robot Trajectory Tracking

Reinforcement Learning-Enhanced Active Disturbance Rejection...

引用

IEEE SSD International Multi-Conference on Systems, Signals and Devices

作者： H. Usama M. Nasir R. Fareh J. Ghommam S. Khadraoui M. Bettayeb Research Institute of Sciences and Engineering University of Sharjah Sharjah UAE Department of Electrical Engineering University of Sharjah Sharjah UAE Department of Electrical and Computer Engineering Sultan Qaboos University Oman Control and Energy Management Laboratory (CEM Lab) Ecole National d'Ingenieur de Sfax Tunisia

ISBN: (数字)9798331542726

ISBN: (纸本)9798331542733

In recent years, the application of robotics has significantly advanced many fields, providing robust performance and efficiency in complex tasks without the need for human intervention. Robot control is a key area of robotics that has received a lot of interest and technological development. Active Disturbance Rejection Controllers (ADRCs) are widely adopted due to several features, including their ability to maintain robust performance and effectively reject disturbances. The manual tuning of controller parameters is time-consuming and highlights the need for automation. This paper presents an innovative approach to tune the parameters of ADRC using Reinforcement Learning (RL) based on Deep Deterministic Policy Gradient (DDPG) for a Differential Drive Mobile Robot (DDMR). The RL agent learns the ideal parameters of the ADRC through real-time, iterative interactions with the simulated environment, improving ADRC's performance without manual tuning. Simulation results demonstrate the effectiveness of the proposed idea in improving the trajectory tracking performance. Combining RL and ADRC provides a promising automated controller tuning solution, opening the door to more intelligent and adaptive robotic systems. 10.0pt.

关键词： Robust control Trajectory tracking Simulation Reinforcement learning Manuals Scattering parameters Mobile robots Artificial intelligence State estimation Tuning

来源：评论

学校读者我要写书评

暂无评论

A Cuboid Geometry Design for MVDC Power Cables for Using in Future All Electric Wide Body Aircraft

A Cuboid Geometry Design for MVDC Power Cables for Using in ...

引用

Conference on electrical Insulation and Dielectric Phenomena, (CEIDP)

作者： Arian Azizi Anoy Saha Mona Ghassemi Department of Electrical and Computer Engineering Zero Emission Realization of Optimized Energy Systems (ZEROES) Laboratory The University of Texas at Dallas Richardson TX USA

Designing high-power-delivery and low-system-mass electric power systems (EPS) is a major goal to achieve the next generation of electrified aircraft. As one of its major components, cables must be redesigned to obtain high-power-density and low-system-mass EPS. Among challenges in designing aircraft cables such as arc and arc tracking, partial discharges (PD), and thermal management, the latter is decisive since the thermal properties of the cable determine its maximum ampacity. The maximum permissible current of a cable depends on radiative and convective heat transfers from its surface to the ambient environment. At the cruising altitude (12.2 km) of wide-body aircraft where the air pressure is 18.8 kPa, the convective heat transfer is greatly reduced which results in a reduction in maximum permissible current. Moreover, both radiative and convective heat transfers depend on the surface area of the cable. One way to increase the heat transfers and compensate for the reduction of convective heat transfer from a limited air pressure is to change the geometry of the cable. The cuboid geometry design provides a larger contact area with the ambient environment for the same cross-section area, so it is expected that the heat transfer will increase compared to conventional cylindrical cables, and in turn, the maximum power carrying capacity of the cable will be larger. Here, the question is whether the hypothesis is true, and if so, how much improvement can be expected. The purpose of this paper is to answer these questions and, for the first time, an MVDC (5 kVdc) high power (1 kA) cuboid shape cable is designed for future AEA to increase the maximum permissible current of the cable.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Geometric Graph Filters and Neural Networks: Limit Properties and Discriminability Trade-offs

arXiv

引用

arXiv 2023年

作者： Wang, Zhiyang Ruiz, Luana Ribeiro, Alejandro Department of Electrical and Systems Engineering University of Pennsylvania PA United States Computer Science & Artificial Intelligence Laboratory MIT MA United States

This paper studies the relationship between a graph neural network (GNN) and a manifold neural network (MNN) when the graph is constructed from a set of points sampled from the manifold, thus encoding geometric information. We consider convolutional MNNs and GNNs where the manifold and the graph convolutions are respectively defined in terms of the Laplace-Beltrami operator and the graph Laplacian. Using the appropriate kernels, we analyze both dense and moderately sparse graphs. We prove non-asymptotic error bounds showing that convolutional filters and neural networks on these graphs converge to convolutional filters and neural networks on the continuous manifold. As a byproduct of this analysis, we observe an important trade-off between the discriminability of graph filters and their ability to approximate the desired behavior of manifold filters. We then discuss how this trade-off is ameliorated in neural networks due to the frequency mixing property of nonlinearities. We further derive a transferability corollary for geometric graphs sampled from the same manifold. We validate our results numerically on a navigation control problem and a point cloud classification task. © 2023, CC0.

关键词： Graph neural networks

来源：评论

学校读者我要写书评

暂无评论

Influence of Solid Dielectric Permittivity on Positive Streamer Characteristics in a Liquid-Solid Composite Dielectric

Influence of Solid Dielectric Permittivity on Positive Strea...

引用

Conference on electrical Insulation and Dielectric Phenomena, (CEIDP)

作者： Mohammad Hamidieh Mona Ghassemi Department of Electrical and Computer Engineering Zero Emission Realization of Optimized Energy Systems (ZEROES) Laboratory The University of Texas at Dallas Richardson TX USA

This paper presents a numerical investigation of positive streamers in a liquid-solid composite dielectric system, specifically in the context of wet-mate DC connectors. The study focuses on the influence of different materials and relative permittivity values of the solid dielectric on streamer behavior. The study employs a non-dimensionalized electric field-dependent molecular ionization streamer model to describe the initiation and propagation of streamers within a needle-sphere electrode system. A 2D-axisymmetric COMSOL model is utilized, where a solid tube-like dielectric is placed near the needle tip in an electrode system filled with transformer oil. The effects of varying relative permittivity values on the electric field distribution, streamer propagation velocity, ionization and attachment rates, and spatiotemporal evolution of charged species (electrons, positive ions, and negative ions) are studied. By analyzing these aspects, the paper aims to enhance the understanding of streamer dynamics and provide valuable information for optimizing the design and performance of equipment utilizing liquid-solid composite dielectric systems.

关键词：

来源：评论

学校读者我要写书评

暂无评论

2D Semiconductor Optoelectronics: Advances, Challenges and Opportunities

2D Semiconductor Optoelectronics: Advances, Challenges and O...

引用

IEEE Nanotechnology Materials and Devices Conference (NMDC)

作者： Ali Javey Department of Electrical Engineering and Computer Sciences University of California Berkeley CA Materials Sciences Division Lawrence Berkeley National Laboratory Berkeley CA

In this talk, I will present recent advancements on understanding and controlling the radiative and non-radiative recombination rates in various 2D semiconductor systems. I will discuss the mechanisms by which non-radiative recombination can be fully suppressed in TMDC monolayers, resulting in near-unity photoluminescence quantum yield at room temperature despite the presence of large defect densities. I will discuss an AC carrier injection mechanism to enable bright light emitting devices using monolayers, overcoming the problem of Schottky contacts. Finally, I will discuss potential applications for black phosphorous (BP) thin films for midwave-IR photo detection and emission. Specifically, the BP based devices are shown to exhibit higher detectivity and luminescence efficiencies over state-of-the-art III-V and II-VI devices in mid-IR, owing to the lower Auger recombination rates and unusually low surface recombination velocity.

关键词：

来源：评论

学校读者我要写书评

暂无评论

Consensus Control of Leader-Following Multi-Agent Systems in Directed Topology With Heterogeneous Disturbances

引用

IEEE/CAA Journal of Automatica Sinica 2021年第2期8卷 423-431页

作者： Qinglai Wei Xin Wang Xiangnan Zhong Naiqi Wu the State Key Laboratory for Management and Control of Complex Systems Institute of AutomationChinese Academy of SciencesBeijing 100190 the School of Artificial Intelligence University of Chinese Academy of SciencesBeijing 100049and also with Qingdao Academy of Intelligent IndustriesQingdao 266109China the Department of Computer and Electrical Engineering and Computer Science Florida Atlantic UniversityBoca Raton FL33431 USA the Institute of Systems Engineering and Collaborative Laboratory for Intelligent Science and SystemsMacao University of Science and TechnologyMacao 999078China

This paper investigates the consensus problem for linear multi-agent systems with the heterogeneous disturbances generated by the Brown *** main contribution is that a control scheme is designed to achieve the dynamic consensus for the multi-agent systems in directed topology interfered by stochastic *** traditional ways,the coupling weights depending on the communication structure are static.A new distributed controller is designed based on Riccati inequalities,while updating the coupling weights associated with the gain matrix by state errors between adjacent *** introducing time-varying coupling weights into this novel control law,the state errors between leader and followers asymptotically converge to the minimum value utilizing the local *** the Lyapunov directed method and It?formula,the stability of the closed-loop system with the proposed control law is *** simulation results conducted by the new and traditional schemes are presented to demonstrate the effectiveness and advantage of the developed control method.

关键词： Consensus control directed topology external disturbance multi-agent(MA)systems

来源：评论

学校读者我要写书评

暂无评论

Short-Term Traffic Flow Prediction Method Based on Spatio-Temporal Characteristics of Complex Road Network

Short-Term Traffic Flow Prediction Method Based on Spatio-Te...

引用

2021 International Conference on Applied Mathematics, Modeling and computer Simulation, AMMCS 2021

作者： Zhu, Huawei Mao, Yihao Zhao, Yiming Shi, Liyuan Shao, Yanli School of Information and Electrical Engineering Zhejiang University City College Hangzhou310015 China Key Laboratory of Complex Systems Modeling and Simulation School of Computer Science and Technology Hangzhou Dianzi University Hangzhou310018 China

ISBN: (纸本)9781643682549

With the development of urbanization, the number of residents' motor vehicles has increased sharply, and traffic congestion problem has become increasingly serious. The construction of Intelligent Traffic System (ITS) has become the main means to alleviate traffic congestion. Short-term traffic flow prediction has guiding significance for residents' travel planning and intelligent management of transportation, and has become one of the research hotspots in intelligent transportation ***, A short-term traffic flow prediction method based on the spatio-temporal characteristics of complex road networks is proposed to further improve the prediction accuracy and reduce the prediction cost. Firstly, a graph convolutional network (GCN) capable of processing non-Euclidean data structures is used to extract the spatial characteristics of traffic flow data. Then, the long and short-term memory (LSTM) neural network is used to process the time characteristics. Finally, the two are combined to realize the effective processing of the spatio-temporal characteristics of traffic flow data. Experimental results on the real traffic flow dataset prove the feasibility and effectiveness of the proposed method, and can provide a basis for intelligent traffic control and smart city construction. © 2022 The authors and IOS Press.

关键词： Traffic congestion

来源：评论

学校读者我要写书评

暂无评论

Exciton complexes and quantum electrodynamics of highly-symmetric single-site-controlled InGaAS quantum dot nanocavities

Exciton complexes and quantum electrodynamics of highly-symm...

引用

CLEO: Science and Innovations, S and I 2022

作者： Huang, Jiahui Liu, Wei Miranda, Alessio Dwir, Benjamin Rudra, Alok Kapon, Eli Wong, Chee Wei Mesoscopic Optics and Quantum Electronics Laboratory Department of Electrical and Computer Engineering University of California Los AngelesCA United States Institute of Physics École Polytechnique Fédérale de Lausanne Lausanne Switzerland

ISBN: (纸本)9781557528209

We demonstrate high-resolution cryogenic photoluminescence measurements on exciton complexes emitting ≈ 1 µm of a single-site-controlled highly-symmetric quantum dot-nanocavity system. Observed fine spectral features uncover the C3ν QD symmetry and cavity quantum electrodynamics. © Optica Publishing Group 2022, © 2022 The Author(s)

关键词： Semiconductor quantum dots

来源：评论

学校读者我要写书评

暂无评论

EffiSegNet: Gastrointestinal Polyp Segmentation through a Pre-Trained EfficientNet-based Network with a Simplified Decoder

EffiSegNet: Gastrointestinal Polyp Segmentation through a Pr...

引用

Annual International Conference of the IEEE engineering in Medicine and Biology Society (EMBC)

作者： Ioannis A. Vezakis Konstantinos Georgas Dimitrios Fotiadis George K. Matsopoulos TECREANDO B.V. Amsterdam The Netherlands Biomedical Engineering Laboratory School of Electrical and Computer Engineering National Technical University of Athens Athens Greece Dept. of Materials Science and Engineering University of Ioannina Ioannina Greece

ISBN: (数字)9798350371499

ISBN: (纸本)9798350371505

This work introduces EffiSegNet, a novel segmentation framework leveraging transfer learning with a pre-trained Convolutional Neural Network (CNN) classifier as its backbone. Deviating from traditional architectures with a symmetric U-shape, EffiSegNet simplifies the decoder and utilizes full-scale feature fusion to minimize computational cost and the number of parameters. We evaluated our model on the gastrointestinal polyp segmentation task using the publicly available Kvasir-SEG dataset, achieving state-of-the-art results. Specifically, the EffiSegNet-B4 network variant achieved an F 1 score of 0.9552, mean Dice (mDice) 0.9483, mean Intersection over Union (mIoU) 0.9056, Precision 0.9679, and Recall 0.9429 with a pre-trained backbone – to the best of our knowledge, the highest reported scores in the literature for this dataset. Additional training from scratch also demonstrated exceptional performance compared to previous work, achieving an F 1 score of 0.9286, mDice 0.9207, mIoU 0.8668, Precision 0.9311 and Recall 0.9262. These results underscore the importance of a well-designed encoder in image segmentation networks and the effectiveness of transfer learning approaches.

关键词： Training Knowledge engineering Image segmentation Transfer learning computer architecture Medical services Gastrointestinal tract Decoding Convolutional neural networks Biomedical imaging

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：