检索结果-内蒙古大学图书馆

2024 IEEE International Conference on Robotics and Biomimetics, ROBIO 2024

作者： Zhang, Xiaoxiong Liu, Junwei Zhang, Wei School of System Design and Intelligent Manufacturing Southern University of Science and Technology Shenzhen Key Laboratory of Control Theory and Intelligent Systems Shenzhen518055 China

ISBN: (纸本)9781665481090

This paper presents a novel learning-based method for a robotic manipulator to achieve liquid pouring across different liquid levels using only visual sensors. Previous works have relied on either online reinforcement learning or imitation learning, which are limited by sim-to-real gaps or data efficiency. In this paper, we propose to combine supervised learning and offline reinforcement learning, utilizing a human demonstration dataset containing only successful pouring at the highest level. Specifically, our approach employs supervised learning for a visual classifier, transforming the visual input into a categorical distribution over liquid levels. The offline reinforcement learning method is applied to a binary-conditioned control policy, which takes a binary signal and the robot's proprioception as inputs to determine the action. The binary signal indicates whether the desired liquid level has been reached based on the target liquid level. Through experiments, our method demonstrates superior effectiveness compared to a state-of-the-art imitation learning algorithm. Moreover, tests in unseen scenarios and multiple liquid level commands verify the generalization and transferability of our approach. © 2024 IEEE.

关键词： Self-supervised learning

来源：评论

学校读者我要写书评

暂无评论

Skeleton-Based Volumetric Parameterizations for Lattice Structures

引用

Computer Modeling in Engineering & Sciences 2023年第4期135卷 687-709页

作者： Long Chen Shuxun Liang Nan Yan Xiangqian Yang Baotong Li School of Mechanical Engineering University of Shanghai for Science and TechnologyShanghai200093China School of Optical-Electrical and Computer Engineering University of Shanghai for Science and TechnologyShanghai200093China Key Laboratory of EducationMinistry for Modern Design&Rotor-Bearing System Xi’an Jiaotong UniversityXi’an710049China

Lattice structures with excellent physical properties have attracted great research *** this paper,a novel volume parametric modeling method based on the skeleton model is proposed for the construction of threedimensional lattice *** skeleton model is divided into three types of *** the corresponding algorithms are utilized to construct diverse types of volume parametric *** unit-cell is assembled with distinct nodes according to the geometric *** final lattice structure is created by the periodic arrangement of *** different types of volume parametric lattice structures are constructed to prove the stability and applicability of the proposed *** quality is assessed in terms of the value of the Jacobian ***,the volume parametric lattice structures are tested with the isogeometric analysis to verify the feasibility of integration of modeling and simulation.

关键词： Lattice structure volumetric parameterization skeleton model node modeling

来源：评论

学校读者我要写书评

暂无评论

VerDT: A Versatile Digital Twins Framework for UAVs-Based Industrial Cyber-Physical systems

引用

IEEE Transactions on Mobile Computing 2025年

作者： Zhou, Longyu Leng, Supeng Quek, Tony Q.S Singapore University of Technology and Design Information System Technology and Design Pillar Singapore School of Information and Communication Engineering Chengdu China

With the development of cyber-physical systems, Digital Twins (DT)-powered network autonomy is emerging to embrace the fifth-generation industrial revolution. In this context, Unmanned Aerial Vehicles (UAVs)-based low-altitude networks are expected to be the engines that drive industrial development. As an attractive industry application, UAVs-based intelligent logistics has been widely investigated to achieve a fully automated distribution manner without the aid of a workforce. However, it is difficult to perform real-time DT implementations due to limited computing resources and the high mobility of UAVs. To address the mentioned problems, we propose a Versatile DT (VerDT) framework operating at the edge. It can enable a double DT cooperation manner with a resource scheduling model and a path planning model for real-time and accurate logistics distributions. The resource scheduling model can implement the integration of computing and communication resources among UAVs for feasible cooperative distribution decisions. With the decisions, the path planning model can imitate to derive positions and velocities of UAVs for low-latency distribution performance with energy saving. Experiment results demonstrate the efficiency of our VerDT framework. Compared to state-of-the-art logistics distribution solutions, our solution reduces the distribution latency by 63.9% while improving the successful distribution ratio by 10.9%. © 2002-2012 IEEE.

关键词： Drones

来源：评论

学校读者我要写书评

暂无评论

Label Recovery and Trajectory designable Network for Transfer Fault Diagnosis of Machines With Incorrect Annotation

引用

IEEE/CAA Journal of Automatica Sinica 2024年第4期11卷 932-945页

作者： Bin Yang Yaguo Lei Xiang Li Naipeng Li Asoke K.Nandi the Key Laboratory of Education Ministry for Modern Design and Rotor-Bearing System the School of Mechanical EngineeringXi’an Jiaotong UniveristyXi’an 710049 the Hunan Provincial Key Labortory of Health Maintenance for Mechanical Equipment Hunan Univeristy of Science and TechnologyXiangtan 411201China the Key Laboratory of Education Ministry for Modern Design and Rotor-Bearing System Xi’an Jiaotong UniveristyXi’an 710049China the Department of Electronic and Electrical Engineering Brunel University LondonUxbridge UB83PHUK

The success of deep transfer learning in fault diagnosis is attributed to the collection of high-quality labeled data from the source ***,in engineering scenarios,achieving such high-quality label annotation is difficult and *** incorrect label annotation produces two negative effects:1)the complex decision boundary of diagnosis models lowers the generalization performance on the target domain,and2)the distribution of target domain samples becomes misaligned with the false-labeled *** overcome these negative effects,this article proposes a solution called the label recovery and trajectory designable network(LRTDN).LRTDN consists of three ***,a residual network with dual classifiers is to learn features from cross-domain ***,an annotation check module is constructed to generate a label anomaly indicator that could modify the abnormal labels of false-labeled samples in the source *** the training of relabeled samples,the complexity of diagnosis model is reduced via semi-supervised ***,the adaptation trajectories are designed for sample distributions across *** ensures that the target domain samples are only adapted with the pure-labeled *** LRTDN is verified by two case studies,in which the diagnosis knowledge of bearings is transferred across different working conditions as well as different yet related *** results show that LRTDN offers a high diagnosis accuracy even in the presence of incorrect annotation.

关键词： Deep transfer learning domain adaptation incorrect label annotation intelligent fault diagnosis rotating machines

来源：评论

学校读者我要写书评

暂无评论

Evaluation of Communication Overhead for Distributed Deep Learning for Local Data Privacy 49

Evaluation of Communication Overhead for Distributed Deep Le...

引用

49th Annual Conference of the IEEE Industrial Electronics Society, IECON 2023

作者： Okuda, Yuma Nishikawa, Akihito Nishi, Hiroaki Graduate School of Science and Technology Keio University Yokohama Japan Keio University Faculty of Science and Technology Department of System Design Yokohama Japan

ISBN: (纸本)9798350331820

In recent years, there has been growing interest in using big data for machine learning applications. However, as data are often distributed across multiple clients, privacy and physical limitations can make it difficult to aggregate and use these data on centralized servers. Federated learning and split neural networks enable machine learning models to be built without requiring the aggregation of distributed data. These methods have received increased attention in the industrial sector as a means of eliminating impediments to privacy preservation in data sharing across various institutions, such as smart factories. One critical aspect of implementing such methods in practice is analyzing the time variation of traffic and estimating the throughput required for operationalizing machine learning. Previous studies have formulated traffic characteristics without considering packet overhead and measured only the total traffic and the throughput per iteration;such an approach is inadequate from a traffic monitoring perspective. To address these issues, this study conducted an empirical investigation of traffic patterns under realistic conditions, with a particular focus on quantifying traffic and latency. The results of this study suggest that due to the dynamic nature of throughput, it is essential to measure traffic using time units that can accommodate the temporal variability of traffic. © 2023 IEEE.

关键词： Data privacy

来源：评论

学校读者我要写书评

暂无评论

Estimation of Indoor Space Temperature Distribution Using Heat Maps 49

Estimation of Indoor Space Temperature Distribution Using He...

引用

49th Annual Conference of the IEEE Industrial Electronics Society, IECON 2023

作者： Hayashi, Emiri Sonoda, Ayu Nishikawa, Akihito Nishi, Hiroaki Graduate School of Science and Technology Keio University Yokohama Japan Keio University Faculty of Science and Technology Department of System Design Yokohama Japan

ISBN: (纸本)9798350331820

In recent years, there has been an increasing demand for accurate measurement of the spatial dispersion of air temperature in air conditioning control for indoor environments. This spatial dispersion of temperature is caused by the unevenness of air conditioning equipment, local temperature gradients due to human activities, and local heat exchange with outdoor air, all of which hinder the provision of a comfortable environment. This issue requires resolution in various fields, including the agricultural sector, where it can negatively impact crop growth. To address this issue, it is necessary to measure air temperature at multiple points, although the number of sensors that can be installed is typically limited, and the temperature at a single point is typically used as a representation of the temperature in the space. Additionally, heat maps using infrared array sensors can acquire temperature distribution in a narrow area;however, they cannot directly determine air temperature as they only measure the surface temperature of an object. Therefore, we propose a deep learning-based method that utilizes both heat maps acquired from infrared array sensors and air temperature data obtained from environmental sensors to estimate air temperature. We evaluate its effectiveness by varying the experimental environment, the model structure, and the resolution and number of heat maps added as input. Furthermore, we assess the usefulness of utilizing a Graphics Processing Unit (GPU), which is a cost-effective solution that can be installed on low-resolution sensors and edge devices. The outcomes indicate that the proposed model can attain an estimation accuracy of under 1 mean squared error (MSE) in approximately 200 seconds, while also exhibiting practical feasibility for deployment on edge devices, including residential abodes and agricultural premises. © 2023 IEEE.

关键词： Temperature distribution

来源：评论

学校读者我要写书评

暂无评论

Lithium-Ion Battery State-of-Charge and State-of-Health Co-Estimation with Feasible Real-Time Implementation 8

Lithium-Ion Battery State-of-Charge and State-of-Health Co-E...

引用

8th CAA International Conference on Vehicular Control and Intelligence, CVCI 2024

作者： Shao, Mingding Chen, Jian Xu, Ke Wang, Yong School of System Design and Intelligent Manufacturing Southern University of Science and Technology Shenzhen China School of Mechanical Engineering Zhejiang University Hangzhou China Wanxiang Al23 Systems Corp. Department of System Engineering Hangzhou China

ISBN: (纸本)9798331504892

The key to improving the efficiency of lithium-ion battery systems and reducing safety incidents is that the battery management system can accurately estimate the state of lithiumion batteries, including the state of charge and the state of health. At present, the commonly used algorithms do not achieve an effective balance between accuracy and complexity, and it is difficult to apply them in practice. Therefore, on the basis of the existing algorithms, this paper proposes to reduce the order of the equivalent circuit model, simplify the observer polynomial, achieve low computational complexity, and use a processor with limited storage to complete the estimations. The simulation and experimental results confirm that the maximum absolute error of the state of charge and state of health of the lithium-ion battery systems fully meet the requirements of practical use. Compared with the reference method, there is no obvious difference in convergence speed and stability, but the run time is relatively faster. The proposed method has a significant simplification effect. © 2024 IEEE.

关键词： State of charge

来源：评论

学校读者我要写书评

暂无评论

LibAMM: Empirical Insights into Approximate Computing for Accelerating Matrix Multiplication 38

LibAMM: Empirical Insights into Approximate Computing for Ac...

引用

38th Conference on Neural Information Processing systems, NeurIPS 2024

作者： Zeng, Xianzhi Jiang, Wenchao Zhang, Shuhao National Engineering Research Center for Big DataTechnology and System Services Computing Technology and System Lab Cluster and Grid Computing Lab School of Computer Science and Technology Huazhong University of Science and Technology Wuhan430074 China Nanyang Technological University Singapore Singapore University of Technology and Design Singapore

Matrix multiplication (MM) is pivotal in fields from deep learning to scientific computing, driving the quest for improved computational efficiency. Accelerating MM encompasses strategies like complexity reduction, parallel and distributed computing, hardware acceleration, and approximate computing techniques, namely AMM algorithms. Amidst growing concerns over the resource demands of large language models (LLMs), AMM has garnered renewed focus. However, understanding the nuances that govern AMM's effectiveness remains incomplete. This study delves into AMM by examining algorithmic strategies, operational specifics, dataset characteristics, and their application in real-world tasks. Through comprehensive testing across diverse datasets and scenarios, we analyze how these factors affect AMM's performance, uncovering that the selection of AMM approaches significantly influences the balance between efficiency and accuracy, with factors like memory access playing a pivotal role. Additionally, dataset attributes are shown to be vital for the success of AMM in applications. Our results advocate for tailored algorithmic approaches and careful strategy selection to enhance AMM's effectiveness. To aid in the practical application and ongoing research of AMM, we introduce LibAMM -a toolkit offering a wide range of AMM algorithms, benchmarks, and tools for experiment management. LibAMM aims to facilitate research and application in AMM, guiding future developments towards more adaptive and context-aware computational solutions. © 2024 Neural information processing systems foundation. All rights reserved.

关键词：

来源：评论

学校读者我要写书评

暂无评论

From Torch to Projector: Fundamental Tradeoff of Integrated Sensing and Communications

IEEE BITS the Information Theory Magazine

引用

IEEE BITS the Information Theory Magazine 2024年第1期4卷 73-90页

作者： Xiong, Yifeng Liu, Fan Wan, Kai Yuan, Weijie Cui, Yuanhao Caire, Giuseppe Beijing University of Posts and Telecommunications School of Information and Electronic Engineering Beijing100876 China Southern University of Science and Technology School of System Design and Intelligent Manufacturing Shenzhen518055 China Huazhong University of Science and Technology School of Electronic Information and Communications Wuhan430074 China Technical University of Berlin Berlin10623 Germany

Sensing and communications (S&C) were historically developed in parallel. In the last decade, they have been evolving from separation to integration, giving rise to the integrated sensing and communication (ISAC) paradigm, that has been recognized as one of 6G key usage scenarios. Despite the plethora of research works dedicated to ISAC signal processing, the fundamental performance limits of S&C remain widely unexplored in ISAC systems. In this tutorial paper, we attempt to summarize the recent research findings in characterizing the performance boundary of ISAC systems and the resulting S&C tradeoff from an information-theoretical viewpoint. We begin with a folkloric "torch metaphor" that depicts the subspace tradeoff between S&C. Then, we elaborate on the fundamental capacity-distortion (C-D) theory, indicating the incompleteness of this metaphor. Toward that end, we further elaborate on the S&C tradeoff by discussing a special case within the C-D framework, namely the Cramér–Rao bound-rate region. In particular, S&C have preference discrepancies over both the signal subspace and the adopted codebook, leading to a "projector metaphor" complementary to the torch analogy. We also present two practical design examples by leveraging the lessons learned from fundamental theories. Finally, we conclude the article by identifying a number of open challenges. © 2021 IEEE.

关键词： Orthogonal frequency division multiplexing

来源：评论

学校读者我要写书评

暂无评论

Research on Sensorless Control of Double High Frequency Injection SynRM Based on Static Shaft Torque Ripple Suppression 2

Research on Sensorless Control of Double High Frequency Inje...

引用

2nd IEEE International Power Electronics and Application Symposium, PEAS 2023

作者： Wang, Jianyuan Jing, Hanghui Li, Yingjie Wang, Haoyu School of Electrical Engineering Xi'an University of Technology Shaanxi Xi'an China School of System Design and Intelligent Manufacturing South University of Science and Technology of China Guangdong Shenzhen China

ISBN: (纸本)9798350313765

When the synchronous reluctance motor runs without position sensor control, there is a strong coupling nonlinear problem of quadrature axis inductance, which affects the precise of rotor position detect and cannot ensure the stability of the motor at low speed range. Control in the low speed range of the motor, a sensorless control method of double high-frequency injection based on static axis torque ripple suppression is proposed. By establishing a nonlinear function model of flux and current of the quadrature axis, the Angle error caused by the cross-coupling of the inductance of the quadrature axis is reduced. The high frequency signal is injected into the d-q axis at the same time to reduce the high frequency torque ripple contained in the high frequency amplitude in the traditional high frequency injection, and the rotor position estimation accuracy and the transient performance of the motor are improved. Finally, the effectiveness of the control scheme is verified by experiments on a 1.5kW synchronous reluctance motor pair drag and load experimental platform. © 2023 IEEE.

关键词： cross saturation double high frequency square wave injection high frequency torque ripple off-line parameter identification synchronous reluctance motor

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：