检索结果-内蒙古大学图书馆

9th International Conference on Science, Technology, engineering and Mathematics, ICONSTEM 2024

作者： Krishna, R. Sathya Barna Sudharani, R. Devi, S. Rukmani Kulkarni, Pramodkumar H. Rohini, G. Sindhu, M. Sathyabama Institute of Science And Technology Department of Computer Science Engineering Tamil Nadu Chennai600119 India Panimalar Engineering College Department of Mathematics Tamil Nadu Poonamallee600123 India Saveetha College of Liberal Arts And Sciences Deemed To Be University SIMATS Department of Computer Science Tamil Nadu Chennai600124 India Dr. D. Y. Patil Institute of Technology Department of Electronics And Telecommunications Maharashtra Pimpri411018 India S.A. Engineering College Department of Electrical And Electronics Engineering Tamil Nadu Thiruverkadu600077 India R.M.K. Engineering College Department of Computer Science And Business Systems Tamil Nadu Thiruvallur601206 India

ISBN: (纸本)9798350365092

Power converter control techniques must evolve to accommodate power system complexity and renewable energy integration. Traditional control methods may be difficult to adapt to these dynamic and uncertain systems. Reinforcement learning (RL) can build optimal control policies in complex and unpredictable situations, making it a promising intelligent control solution in this scenario. This research examines how reinforcement learning can control power converters in energy systems. Advanced reinforcement learning methods, notably deep reinforcement learning, are used to improve power converter efficiency in varied operational scenarios. This research examines the integration of reinforcement learning (RL)-based controllers into power converter systems, focusing on their autonomous learning and adaptation. The explored method uses Reinforcement Learning (RL) flexibility to address power system time-varying conditions, non-linearities, and uncertainties. This research considers practical concerns such RL-based control's effects on real-time execution, processing requirements, and safety. The report concludes with suggestions for future research including expanding RL-based intelligent control systems to power system applications. Reinforcement learning keywords include stability, efficiency, reliability, renewable energy, deep reinforcement learning, power converters, intelligent control, and energy systems. © 2024 IEEE.

关键词： Power converters

来源：评论

学校读者我要写书评

暂无评论

TeMPO: Efficient Time-Multiplexed Dynamic Photonic Tensor Core for Edge AI with Compact Slow-Light Electro-Optic Modulator

arXiv

引用

arXiv 2024年

作者： Zhang, Meng Yin, Dennis Gangi, Nicholas Begović, Amir Chen, Alexander Huang, Zhaoran Rena Gu, Jiaqi Department of Electrical Computer and Systems Engineering Rensselaer Polytechnic Institute TroyNY12180 United States School of Electrical Computer and Energy Engineering Arizona State University TempeAZ85287 United States

Electronic-photonic computing systems offer immense potential in energy-efficient artificial intelligence (AI) acceleration tasks due to the superior computing speed and efficiency of optics, especially for real-time, low-energy deep neural network (DNN) inference tasks on resource-restricted edge platforms. However, current optical neural accelerators based on foundry-available devices and conventional system architecture still encounter a performance gap compared to highly customized electronic counterparts. To bridge the performance gap due to lack of domain specialization, we present a time-multiplexed dynamic photonic tensor accelerator, dubbed TeMPO, with cross-layer device/circuit/architecture customization. At the device level, we present foundry-compatible, customized photonic devices, including a slow-light electro-optic modulator with experimental demonstration, optical splitters, and phase shifters that significantly reduce the footprint and power in input encoding and dot-product calculation. At the circuit level, partial products are hierarchically accumulated via parallel photocurrent aggregation, lightweight capacitive temporal integration, and sequential digital summation, considerably relieving the analog-to-digital conversion bottleneck. We also employ a multi-tile, multi-core architecture to maximize hardware sharing for higher efficiency. Across diverse edge AI workloads, TeMPO delivers digital-comparable task accuracy with superior quantization/noise tolerance. We achieve a 368.6 TOPS peak performance, 22.3 TOPS/W energy efficiency, and 1.2 TOPS/mm2 compute density, pushing the Pareto frontier in edge AI hardware. This work signifies the power of cross-layer co-design and domain-specific customization, paving the way for future electronic-photonic accelerators with even greater performance and efficiency. Copyright © 2024, The Authors. All rights reserved.

关键词： Memory architecture

来源：评论

学校读者我要写书评

暂无评论

An Effective Image-Based Tomato Leaf Disease Segmentation Method Using MC-UNet

引用

Plant Phenomics 2023年第2期5卷 268-284页

作者： Yubao Deng Haoran Xi Guoxiong Zhou Aibin Chen Yanfeng Wang Liujun Li Yahui Hu College of Computer&Information Engineering Central South University of Forestry and TechnologyChangsha410004HunanChina College of Mechanical&Electrical Engineering Central South University of Forestry and TechnologyChangsha410004HunanChina National University of Defense Technology 410015ChangshaHunanChina Department of Soil and Water Systems University of IdahoMoscowID83844USA Plant Protection Research Institute Academy of Agricultural Sciences410125ChangshaHunanChina

Tomato disease control is an urgent requirement in the field of intellectual agriculture,and one of the keys to it is quantitative identification and precise segmentation of tomato leaf *** diseased areas on tomato leaves are tiny and may go unnoticed during *** edge also makes the segmentation accuracy *** on UNet,we propose an effective image-based tomato leaf disease segmentation method called Cross-layer Attention Fusion Mechanism combined with Multi-scale Convolution Module(MC-UNet).First,a Multi-scale Convolution Module is *** module obtains multiscale information about tomato disease by employing 3 convolution kernels of different sizes,and it highlights the edge feature information of tomato disease using the Squeeze-and-Excitation ***,a Cross-layer Attention Fusion Mechanism is *** mechanism highlights tomato leaf disease locations via gating structure and fusion ***,we employ SoftPool rather than MaxPool to retain valid information on tomato ***,we use the SeLU function appropriately to avoid network neuron *** compared MC-UNet to the existing segmentation network on our self-built tomato leaf disease segmentation dataset and MC-UNet achieved 91.32%accuracy and 6.67M *** method achieves good results for tomato leaf disease segmentation,which demonstrates the effectiveness of the proposed methods.

关键词： network precise sizes

来源：评论

学校读者我要写书评

暂无评论

ACL-QL: Adaptive Conservative Level in Q-Learning for Offline Reinforcement Learning

引用

IEEE Transactions on Neural Networks and Learning systems 2024年第6期36卷 11399-11413页

作者： Kun Wu Yinuo Zhao Zhiyuan Xu Zhengping Che Chengxiang Yin Chi Harold Liu Feifei Feng Jian Tang Department of Electrical Engineering and Computer Science Syracuse University Syracuse NY USA Beijing Institute of Technology Beijing China Beijing Innovation Center of Humanoid Robotics Beijing China Midea Group Shanghai China

Offline reinforcement learning (RL), which operates solely on static datasets without further interactions with the environment, provides an appealing alternative to learning a safe and promising control policy. The prevailing methods typically learn a conservative policy to mitigate the problem of Q-value overestimation, but it is prone to overdo it, leading to an overly conservative policy. Moreover, they optimize all samples equally with fixed constraints, lacking the nuanced ability to control conservative levels in a fine-grained manner. Consequently, this limitation results in a performance decline. To address the above two challenges in a united way, we propose a framework, adaptive conservative level in Q-learning (ACL-QL), which limits the Q-values in a mild range and enables adaptive control on the conservative level over each state-action pair, i.e., lifting the Q-values more for good transitions and less for bad transitions. We theoretically analyze the conditions under which the conservative level of the learned Q-function can be limited in a mild range and how to optimize each transition adaptively. Motivated by the theoretical analysis, we propose a novel algorithm, ACL-QL, which uses two learnable adaptive weight functions to control the conservative level over each transition. Subsequently, we design a monotonicity loss and surrogate losses to train the adaptive weight functions, Q-function, and policy network alternatively. We evaluate ACL-QL on the commonly used datasets for deep data-driven reinforcement learning (D4RL) benchmark and conduct extensive ablation studies to illustrate the effectiveness and state-of-the-art performance compared with existing offline DRL baselines.

关键词： Adaptive systems Reinforcement learning Q-learning Weight measurement Loss measurement Training Propagation losses Navigation Learning systems Imitation learning

来源：评论

学校读者我要写书评

暂无评论

A Voxel-Enabled Robotic Assistant for Omnidirectional Conveyance

A Voxel-Enabled Robotic Assistant for Omnidirectional Convey...

引用

IEEE/RSJ International Conference on Intelligent Robots and systems (IROS)

作者： Michael Angelo Carvajal Katiso Mabulu Muneer Lalji James Flanagan Rui Luo Samuel Hibbard Tanav Chinthapatla Rohan Bettadpur Salah Bazzi Mark Zolotas Kristian Kloeckl Taşkın Padır Institute for Experiential Robotics Northeastern University Boston MA USA Department of Art + Design School of Architecture Northeastern University Electrical and Computer Engineering Northeastern University

ISBN: (数字)9798350377705

ISBN: (纸本)9798350377712

Conventional bidirectional conveyance platforms use a flat translating belt or a series of spinning wheels or rollers to apply a shear force to payloads to move them. Wheel/roller-based conveyors in particular cannot double as a worktop when idle, do not support collision-free multi-object manipulation by default, and are not optimized to move objects that are either slippery or pliable—let alone both. This paper introduces a Voxel-Enabled Robotic Assistant (VERA), a network of intelligent table "partitions" whose topologically dynamic worktops enable omnidirectional conveyance; each partition is composed of a 2D array of "quadrants," axisymmetric modules that can be hot-swapped for maintenance or repairs; each quadrant contains a 2D array of "cells," unitary robotic submodules; each cell houses an independently controllable "voxel," the motorized rotary element that conveys an overhead object. The efficacy of a VERA prototype was determined by evaluating waypoint error as a range of payloads were maneuvered between trajectory waypoints. By conveying both pliable and rigid payloads having slippery textures, the faceted voxels outperformed those augmented to mimic the circular-profiled wheels/rollers of competitor systems. VERA also successfully performed collision-free multi-object planar manipulations planned by its pathfinding algorithm. In light of these results, VERA emerges as a promising material handling platform for use in "Future of Work" settings as the need for multi-purpose collaborative industrial robots continues to grow.

关键词： Visualization Translation Service robots Prototypes Wheels Trajectory Mobile robots Collision avoidance Robots Payloads

来源：评论

学校读者我要写书评

暂无评论

Portable Automatic Sensing System for Sustainable Precision Farm

Renewable Energy and Power Quality Journal

引用

Renewable Energy and Power Quality Journal 2022年 20卷 222-227页

作者： Galvão, J. Fernandes, J. Pereira, D. Galvão, M. Neves, F. Department of Electrical Engineering Polytechnic of Leiria Leiria2411-901 Portugal R&D Unit Institute for Systems Engineering and Computers INESC Coimbra Portugal ISTécnico/Lisbon Technical University Lisboa1049-001 Portugal Department of Computer Science Polytechnic of Leiria Leiria2411-901 Portugal

Taking into account the world´s growing population and climate change, there is a need to decarbonize production processes and, in particular, intensive cultures carried out in greenhouses. There is a need to boost the energy and water efficiency of these cyber-physical systems, taking into account human consumption or floriculture, as well as European Community (EEC) and United Nations agreements and pledges, such as the European Green Deal (EGD) or the REPowerEU Plan [1,2] t is critical to develop systems that use integrated models to control these sorts of intense production, collect physical variable factors more efficiently, and achieve improved and zero greenhouse gas emissions (GHG). Finally, it is crucial for a digitalization in these intensive crops, in precision farm or greenhouses. The main objective of this work focuses on the conception of a portable automatic sensing system (PASS) for pH, electroconductivity and water flow, to control irrigation water quality in precision farm/greenhouses. The data collected by these sensors are made available in a mobile app for consultation by the operator. Another objective will be the local production of electrical and thermal energy from renewable sources (solar and biomass), hybrid systems and the use of cogeneration to supply energy to the greenhouse. A sustainable product policy also has the potential to reduce waste significantly. © 2022, European Association for the Development of Renewable Energy, Environment and Power Quality (EA4EPQ). All rights reserved.

关键词： Sustainable development

来源：评论

学校读者我要写书评

暂无评论

Risk-Averse Learning with Delayed feedback

arXiv

引用

arXiv 2024年

作者： Wang, Siyi Wang, Zifan Johansson, Karl H. Hirche, Sandra Department of Electrical and Computer Engineering Technical University of Munich Munich80333 Germany Division of Decision and Control Systems School of Electrical Engineering and Computer Science KTH Royal Institute of Technology Stockholm10044 Sweden

In real-world scenarios, the impacts of decisions may not manifest immediately. Taking these delays into account facilitates accurate assessment and management of risk in real-world environments, thereby ensuring the efficacy of strategies. In this paper, we investigate risk-averse learning using Conditional Value at Risk (CVaR) as risk measure, while incorporating delayed feedback with unknown but bounded delays. We develop two risk-averse learning algorithms that rely on one-point and two-point zeroth-order optimization approaches, respectively. The regret achieved by the algorithms is analyzed in terms of the cumulative delay and the number of total samplings. The results suggest that the two-point risk-averse learning achieves a smaller regret bound than the one-point algorithm. Furthermore, the one-point risk-averse learning algorithm attains sublinear regret under certain delay conditions, and the two-point risk-averse learning algorithm achieves sublinear regret with minimal restrictions on the delay. We provide numerical experiments on a dynamic pricing problem to demonstrate the performance of the proposed algorithms. © 2024, CC0.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Multi-Layer Network Formation Through HAPS Base Station and Transmissive RIS-Equipped UAV

Multi-Layer Network Formation Through HAPS Base Station and ...

引用

2025 IEEE Wireless Communications and Networking Conference, WCNC 2025

作者： Khennoufa, Faical Abdellatif, Khelil Yanikomeroglu, Halim Ozturk, Metin Elganimi, Taissir Kara, Ferdi Rabie, Khaled El-Oued University Lgeere Lab Department of Electrical Engineering El-Oued Algeria Carleton University Ntn Lab Department of Systems and Computer Engineering Ottawa Canada Ankara Yildirim Beyazit University Electrical and Electronics Engineering Ankara Turkey University of Tripoli Department of Electrical and Electronic Engineering Libya Kth Royal Institute of Technology Division of Communications Systems Sweden Zonguldak Bulent Ecevit University Department of Computer Engineering Zonguldak Turkey Department of Computer Engineering Center for Communication Systems and Sensing Dhahran Saudi Arabia

ISBN: (纸本)9798350368369

In order to bolster future wireless networks, there has been a great deal of interest in non-terrestrial networks, especially aerial platforms including high-altitude platform stations (HAPS) and uncrewed aerial vehicles (UAVs). These platforms can integrate advanced technologies such as reconfigurable intelligent surfaces (RIS) and non-orthogonal multiple access (NOMA). In this regard, this paper proposes a multi-layer network architecture consisting of HAPS and UAV, where the former acts as a HAPS super macro base station (HAPS-SMBS), while the latter serves as a relay node for the ground Internet of Things (IoT) devices. The UAV is equipped with active transmissive RIS, which is a novel technology with promising benefits. We also utilize multiple-input single-output (MISO) technology, i.e., multiple antennas at the HAPS-SMBS and a single antenna at the IoT devices. Additionally, we consider NOMA as the multiple access technology as well as the existence of hardware impairments as a practical limitation. We compare the proposed system model with various scenarios, all involving the HAPS-SMBS and RIS-equipped UAV relay combination, but with different types of RIS, antenna configurations, and access technologies. Sum rate and energy efficiency are used as performance metrics, and the findings demonstrate that, in comparison to all benchmarks, the proposed system yields significant performance gains. Moreover, hardware impairment limits the system performance at high transmit power levels. © 2025 IEEE.

关键词： Enterprise resource planning

来源：评论

学校读者我要写书评

暂无评论

Estimation of External Force acting on Underwater Robots

Estimation of External Force acting on Underwater Robots

引用

IEEE International Conference on Automation Science and engineering (CASE)

作者： Srinivasan Lakshminarayanan Daniel Duecker Andriy Sarabakha Amartya Ganguly Leila Takayama Sami Haddadin Munich Institute of Robotics and Machine Intelligence (MIRMI) TU Munich Germany Department of Electrical and Computer Engineering Aarhus University Denmark Department of Computational Media University of California USA

ISBN: (数字)9798350358513

ISBN: (纸本)9798350358520

In the context of robotics, accurately estimating external forces is essential. These estimations play a crucial role in collision detection, motion planning, robot control, disturbance compensation, environmental interactions, etc. Given the importance of external force estimation, We introduce a framework to estimate the forces acting on underwater robots. By leveraging dynamic models of the underwater remotely operated vehicle (ROV), our framework predicts and estimates forces based on velocity and acceleration data. We conducted real-world experiments on BlueROV2 underwater vehicle to test and verify the efficacy of our proposed framework. We determined that our framework can predict the external forces with an MAE of 0.8 N.

关键词： Underwater structures Autonomous underwater vehicles Force measurement Remotely guided vehicles Force Robot control Estimation Mathematical models Vehicle dynamics Collision avoidance

来源：评论

学校读者我要写书评

暂无评论

Generalized Continuous-Time Models for Nesterov’s Accelerated Gradient Methods

arXiv

引用

arXiv 2024年

作者： Park, Chanwoong Cho, Youngchae Yang, Insoon Department of Electrical and Computer Engineering Seoul National University Seoul08826 Korea Republic of Automation and Systems Research Institute Seoul National University Seoul08826 Korea Republic of Department of Electrical and Computer Engineering Automation and Systems Research Institute Seoul National University Seoul08826 Korea Republic of

Recent research has indicated a substantial rise in interest in understanding Nesterov’s accelerated gradient methods via their continuous-time models. However, most existing studies focus on specific classes of Nesterov’s methods, which hinders the attainment of an in-depth understanding and a unified perspective. To address this deficit, we present generalized continuous-time models that cover a broad range of Nesterov’s methods, including those previously studied under existing continuous-time frameworks. Our key contributions are as follows. First, we identify the convergence rates of the generalized models, eliminating the need to determine the convergence rate for any specific continuous-time model derived from them. Second, we show that six existing continuous-time models are special cases of our generalized models, thereby positioning our framework as a unifying tool for analyzing and understanding these models. Third, we design a restart scheme for Nesterov’s methods based on our generalized models and show that it ensures a monotonic decrease in objective function values. Owing to the broad applicability of our models, this scheme can be used to a broader class of Nesterov’s methods compared to the original restart scheme. Fourth, we uncover a connection between our generalized models and gradient flow in continuous time, showing that the accelerated convergence rates of our generalized models can be attributed to a time reparametrization in gradient flow. Numerical experiment results are provided to support our theoretical analyses and results. © 2024, CC BY-NC-ND.

关键词： Continuous time systems

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：