检索结果-内蒙古大学图书馆

IEEE Transactions on Artificial Intelligence 2024年第7期5卷 3607-3623页

作者： Mandal, Lakshmi Diddigi, Raghuram Bharadwaj Bhatnagar, Shalabh Indian Institute of Science Department of Computer Science and Automation Karnataka Bangalore560012 India International Institute of Information Technology Karnataka Bangalore560100 India

Reinforcement learning (RL) algorithms combined with deep learning architectures have achieved tremendous success in many practical applications. However, the policies obtained by many deep reinforcement learning (DRL) algorithms are seen to suffer from high variance making them less useful in safety-critical applications. In general, it is desirable to have algorithms that give a low iterate-variance while providing a high long-term reward. In this work, we consider the actor-critic (AC) paradigm, where the critic is responsible for evaluating the policy while the feedback from the critic is used by the actor for updating the policy. The updates of both the critic and the actor in the standard AC procedure are run concurrently until convergence. It has been previously observed that updating the actor once after every L>1 steps of the critic reduces the iterate variance. In this article, we address the question of what optimal L-value to use in the recursions and propose a data-driven L-update rule that runs concurrently with the AC algorithm with the objective being to minimize the variance of the infinite horizon discounted reward. This update is based on a random search (discrete) parameter optimization procedure that incorporates smoothed functional (SF) estimates. We prove the convergence of our proposed multitimescale scheme to the optimal L and optimal policy pair. Subsequently, through numerical evaluations on benchmark RL tasks, we demonstrate the advantages of our proposed algorithm over multiple state-of-the-art algorithms in the literature. © 2020 IEEE.

关键词： Optimization

来源：评论

学校读者我要写书评

暂无评论

Coordinated Frequency Control for Isolated Power Systems with High Penetration of DFIG-Based Wind Power

引用

CSEE Journal of Power and Energy Systems 2024年第4期10卷 1399-1414页

作者： Xin Ding Wei Lin Jian Xu Yuanzhang Sun Liangzhong Yao Beilin Mao School of Electrical Engineering and Automation Wuhan UniversityWuhan 430072China Department of Electrical Engineering and Computer Science Case Western Reserve UniversityClevelandOH 44106 USA

This paper proposes a coordinated frequency control scheme for emergency frequency regulation of isolated power systems with a high penetration of wind *** proposed frequency control strategy is based on the novel nonlinear regulator theory,which takes advantage of nonlinearity of doubly fed induction generators(DFIGs)and generators to regulate the frequency of the power *** deviations and power imbalances are used to design nonlinear feedback controllers that achieve the reserve power distribution between generators and DFIGs,in various wind speed *** effectiveness and dynamic performance of the proposed nonlinear coordinated frequency control method are validated through simulations in an actual isolated power grid.

关键词： Frequency control isolated power systems nonlinear feedback nonlinear regulator theory wind power

来源：评论

学校读者我要写书评

暂无评论

Autonomous embodied navigation task generation from natural language dialogues

引用

science China(Information sciences) 2025年第5期68卷 119-132页

作者： Haifeng XU Yongchang LI Lumeng MA Chunwen LI Yanzhi DONG Xiaohu YUAN Huaping LIU Department of Automation Tsinghua University School of Physics and Electronic Information Yantai University School of Electrical and Electronic Engineering Shanghai Institute of Technology Department of Computer Science and Technology Tsinghua University

Robots are increasingly being deployed in densely populated environments, such as homes, hotels, and office buildings, where they rely on explicit instructions from humans to perform tasks. However, complex tasks often require multiple instructions and prolonged monitoring, which can be time-consuming and demanding for users. Despite this, there is limited research on enabling robots to autonomously generate tasks based on real-life scenarios. Advanced intelligence necessitates robots to autonomously observe and analyze their environment and then generate tasks autonomously to fulfill human requirements without explicit commands. To address this gap, we propose the autonomous generation of navigation tasks using natural language dialogues. Specifically, a robot autonomously generates tasks by analyzing dialogues involving multiple persons in a real office environment to facilitate the completion of item transportation between various *** propose the leveraging of a large language model(LLM) through chain-of-thought prompting to generate a navigation sequence for a robot from dialogues. We also construct a benchmark dataset consisting of 625 multiperson dialogues using the generation capability of LLMs. Evaluation results and real-world experiments in an office building demonstrate the effectiveness of the proposed method.

关键词： proactive robot robot navigation service robot large language model

来源：评论

学校读者我要写书评

暂无评论

Privacy preserving federated learning with convolutional variational bottlenecks

引用

Cybersecurity 2025年第1期8卷 1-22页

作者： Scheliga, Daniel Mäder, Patrick Seeland, Marco Department of Computer Science and Automation Technische Universität Ilmenau Helmholtzplatz 5 Thuringia Ilmenau98693 Germany Faculty of Biological Sciences Friedrich Schiller University Jena Germany

Gradient Inversion (GI) attacks are a ubiquitous threat in Federated Learning as they exploit gradient leakage to reconstruct supposedly private training data. Recent work has proposed to prevent gradient leakage without loss of model utility by incorporating a PRivacy EnhanCing mODulE (PRECODE) based on variational modeling. Without further analysis, it was shown that PRECODE successfully protects against GI attacks. In this paper, we make multiple contributions. First, we investigate the effect of PRECODE on GI attacks to reveal its underlying working principle. We show that variational modeling introduces stochasticity into the gradients of PRECODE and the subsequent layers in a neural network. The stochastic gradients of these layers prevent iterative GI attacks from converging. Second, we formulate an attack that disables the privacy preserving effect of PRECODE by purposefully omitting stochastic gradients during attack optimization. To preserve the privacy preserving effect of PRECODE, our analysis reveals that variational modeling must be placed early in the network. However, early placement of PRECODE is typically not feasible due to reduced model utility and the exploding number of additional model parameters. Therefore, as a third contribution, we propose a novel privacy module—the Convolutional Variational Bottleneck (CVB)—that can be placed early in a neural network without suffering from these drawbacks. We conduct an extensive empirical study on three seminal model architectures and six image classification datasets. We find that all architectures are susceptible to GI attacks, which can be prevented by our proposed CVB. Compared to PRECODE, we show that our novel privacy module requires fewer trainable parameters, and thus computational and communication costs, to effectively preserve privacy. © The Author(s) 2025.

关键词： Federated learning

来源：评论

学校读者我要写书评

暂无评论

eHealth Mobile Application Using Fitbit Smartwatch 1

引用

8th International Conference on Advancements of Medicine and Health Care through Technology, MEDITECH 2022

作者： Pop, Adela Fanca, Alexandra Valean, Honoriu Gota, Dan-Ioan Stan, Ovidiu Roman, Marius Nicolae Clitan, Iulia Muresan, Vlad Faculty of Automation and Computer Science Department of Automation Technical University of Cluj Napoca Cluj Napoca Romania Faculty of Electrical Engineering Electrical Engineering and Measurements Department Technical University of Cluj Napoca Cluj Napoca Romania

ISBN: (数字)9783031511202

ISBN: (纸本)9783031511196

Health is one of the most precious aspects of every person's life. Whether we're talking about a well-functioning heart, nervous system, respiratory system, or simply body weight, each of these branches contributes to a person's overall health. It is important to monitor certain body functions and characteristics as often as possible so that we can prevent or, in unfortunate situations, treat health problems early. The developed mobile application facilitates the monitoring of heart rate, ideal weight, basal metabolic rate, and body mass index. The data used by the mobile app is collected by the FITBIT smartwatch. Based on the data, heart rate statistics and nutritional calculation results (body mass index, basal metabolic rate, and ideal body weight) are generated according to the user's gender, age, weight, and height. © 2024, The Author(s), under exclusive license to Springer Nature Switzerland AG.

关键词： Metabolism

来源：评论

学校读者我要写书评

暂无评论

Control of ROV Thrusters via Discrete Modal State Controller

Control of ROV Thrusters via Discrete Modal State Controller

引用

2023 International Conference Automatics and Informatics, ICAI 2023

作者： Nikolov, Nikola N. Grigorov, Ivan V. Chterev, Konstantin I. Technical University of Varna Faculty of Computer Science and Automation Department of Automation Varna Bulgaria Black Sea Research Fund Varna Bulgaria

ISBN: (纸本)9798350312911

Most of the underwater activities are carried out with a Remotely Operated Vehicle (ROV). ROVs are driven by thrusters, and controlling the movement of the ROV underwater is related to continuous regulation of the speed of the electric motors driving its thrusters. This paper presents one variant of ROV propulsion control using a discrete modal state controller. A mathematical description in the state space has been made for one of the most commonly used electric motors. The operation algorithm of the discrete modal state controller to calculate the feedback vector was developed and simulation modeling was done in MATLAB/Simulink. © 2023 IEEE.

关键词： MATLAB

来源：评论

学校读者我要写书评

暂无评论

MangoLeafXNet: An Explainable Deep Learning Model for Accurate Mango Leaf Disease Classification

引用

IEEE Access 2025年 13卷 93977-94008页

作者： Rayed, Md Eshmam Jim, Jamin Rahman Islam, Md Juniadul Mridha, M.F. Kabir, Md Mohsin Hossen, Md. Jakir American International University-Bangladesh Department of Computer Science and Engineering Dhaka1229 Bangladesh Mälardalens University Division of Computer Science and Software Engineering Västerås722 20 Sweden Multimedia University Faculty of Engineering and Technology Department of Robotics and Automation Melaka Malaysia

Addressing the global challenge of ensuring a consistent and abundant supply of fresh fruit, particularly in the context of fruit crops, is hindered by the prevalence of plant diseases. These diseases directly impact the quality of fruits, leading to a decline in overall agricultural production. Mango leaf diseases pose significant threats to global mango production, necessitating accurate and efficient classification techniques for timely disease management. Our study focuses on introducing MangoLeafXNet, a customized Convolutional Neural Network (CNN) architecture specifically tailored for the classification of mango leaf diseases, along with a healthy class. Our proposed model comprises six layers optimized to capture intricate disease patterns, demonstrating superior performance compared with prevalent pre-trained models. The model is trained and evaluated on three publicly available datasets: MangoLeafBD (4000 images across 8 classes), MangoPest (16 pest classes including healthy leaves), and MLDID (3000 high-resolution images across 5 classes). Our model demonstrated exceptional classification performance, attaining 99.8% accuracy, 99.62% recall, 99.5% precision, and an F1-score of 99.56%. Further validation on the MangoPest dataset and the Mango Leaf Disease Identification Dataset (MLDID) resulted in accuracies of 96.31% and 96.33%, respectively, confirming the robustness and adaptability of MangoLeafXNet across different datasets. Additionally, we incorporate Explainable AI techniques, including GRAD-CAM, Saliency Map, and LIME to enhance the interpretability of our model. We deployed Gradio web interface to create an interactive interface that allows users to upload images of mango leaves and get real-time classification and validation results along with confidence scores. This contribution not only advances the state-of-the-art in mango leaf disease classification but also offers promising prospects for real-time disease diagnosis and precision agriculture

关键词： Fruits

来源：评论

学校读者我要写书评

暂无评论

University - Research - Farm Cooperation to Identify the Halyomorpha Halys Invasion in Orchards 14

University - Research - Farm Cooperation to Identify the Hal...

引用

14th IEEE Global Engineering Education Conference, EDUCON 2023

作者： Popescu, Dan Ichim, Loretta University POLITEHNICA of Bucharest Faculty of Automation and Computer Science Bucharest Romania

ISBN: (纸本)9798350399431

Nowadays, multidisciplinary research is found everywhere, especially due to the accelerated development of computer science and information and communication technology in all economic and social fields. In many universities, including the POLITEHNICA University in Bucharest, research as a subject is considered multidisciplinary through case studies and concrete applications. The paper presents a multidisciplinary project implemented under ERA-NET COFOUND, ICT-AGRI-FOOD 2019 Joint Call. This project implies multiple collaborations inside and between different groups: disciplines (computer science-image processing-artificial intelligence-robotics-entomology-agriculture), domains (universities-research-farms), countries, and teams (teachers-researchers-students-farmers). The resulting impact of this collaboration is also multiple: educational, economic, and social. From an educational point of view, the results consisted of course chapters, themes for undergraduate, master, and doctoral degrees, and papers in important international conferences or journals made by teams of students, researchers, and professors. © 2023 IEEE.

关键词： Farms

来源：评论

学校读者我要写书评

暂无评论

Interference Suppression and Jitter Elimination Ability-Based Adaption Tracking Guidance for Robotic Fishes

引用

IEEE/CAA Journal of Automatica Sinica 2025年第1期12卷 126-137页

作者： Dongfang Li Jie Huang Rob Law Xin Xu Limin Zhu Edmond Q.Wu IEEE the School of Electrical Engineering and Automation Fuzhou University the Key Laboratory of System Control and Information Processing Ministry of Education the School of Electrical Engineering and Automation Fuzhou University 5G+Industrial Internet Institute Fuzhou University the University of Macau the College of Mechatronics and Automation National University of Defense Technology the School of Mechanical Engineering Shanghai Jiao Tong University the Department of Computer Science and Engineering Shanghai Jiao Tong University

This work presents an adaptive tracking guidance method for robotic fishes. The scheme enables robots to suppress external interference and eliminate motion jitter. An adaptive integral surge line-of-sight guidance rule is designed to eliminate dynamics interference and sideslip issues. Limited-time yaw and surge speed observers are reported to fit disturbance variables in the model. The approximation values can compensate for the system's control input and improve the robots' tracking ***, this work develops a terminal sliding mode controller and third-order differential processor to determine the rotational torque and reduce the robots' run jitter. Then, Lyapunov's theory proves the uniform ultimate boundedness of the proposed method. Simulation and physical experiments confirm that the technology improves the tracking error convergence speed and stability of robotic fishes.

关键词： Adaptive integral surge line-of-sight approximation robotic fish speed observer

来源：评论

学校读者我要写书评

暂无评论

MOLE - a Mobile Lab Extension Kit for GOLDi Remote Labs 6

MOLE - a Mobile Lab Extension Kit for GOLDi Remote Labs

引用

6th Experiment at International Conference, exp.at 2023

作者： Henke, Karsten Nau, Johannes Ilmenau University of Technology Faculty of Computer Science and Automation Ilmenau Germany

ISBN: (纸本)9798350314403

Our vision is, that an online lab will no longer be seen as a collection of monolithically constructed experiments but as a collection of laboratory devices that communicate with each other. Nowadays, many interested learners have their own hardware at home. Therefore, an extension to the GOLDi remote lab, we provide learners with an interface unit, which in turn means: they can easily connect to this interface with their control units and thus have the possibility to control these complex hardware models (from home). © 2023 IEEE.

关键词： E-learning

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：