检索结果-内蒙古大学图书馆

9th International Conference on Mechanical engineering and robotics Research, ICMERR 2025

作者： Fernandez, Manuel J. Franceschini, Riccardo Becerra, Julian Cayero Fumagalli, Matteo Technical University of Denmark Department of Electrical Engineering Automation and Control Kongens Lyngby2800 Denmark Robotics and Automation Unit Eurecat Technology Center of Catalunya Cerdanyola Del Vallès Barcelona Spain

ISBN: (纸本)9798331509262

As aerial robots evolve to perform complex tasks such as manipulation and visual inspection, maintaining the pilot at the center of decision-making is critical for mission safety and leveraging human expertise. A key enabler of this collaboration is reliable video streaming between the robot and the operator. This paper presents a comparative analysis of four video streaming protocols RTP (Real-time Transport Protocol), RTSP (Real-Time Streaming Protocol), SRT (Secure Reliable Transport), and ROSTCP (ROS Transport control Protocol) based on extensive field experiments. The evaluation focuses on CPU usage, network throughput, latency, streaming quality, and packet integrity across varying distances. Tools such as iperf, ping, GStreamer, and Wireshark were employed for detailed performance measurements. Results reveal significant differences: SRT and RTP deliver superior network stability and resilience to packet loss, while RTSP and ROSTCP show advantages in CPU efficiency. These findings provide actionable insights for optimizing video streaming in aerial robotics, emphasizing the trade-offs and advantages of each protocol in real-world scenarios. © 2025 IEEE.

关键词： Social robots

来源：评论

学校读者我要写书评

暂无评论

Task Load Based Persistent Coverage control for Large-Scale Domains Using Multi-Agent Systems with Anisotropic Sensors

Task Load Based Persistent Coverage Control for Large-Scale ...

引用

IEEE Southeastcon

作者： Kyle Odoy Erick J. Rodríguez-Seda Department of Weapons Robotics and Control Engineering United States Naval Academy Annapolis MD

ISBN: (数字)9798331504847

ISBN: (纸本)9798331504854

Multi-agent coverage control refers to the task of coordinating autonomous robots to effectively survey and gather information across a task-domain. This paper introduces a novel approach that addresses agents equipped with anisotropic sensors and constrained by nonholonomic movement, aiming to ensure persistent coverage and maintain awareness over a large-scale domain. The approach employs redefined Voronoi partitions that leverage each agent's local coverage domain based on its relative task load. The task load herein refers to the level of awareness that each agent has about the domain and the size of its area of dominance, all indirectly encoded in the Voronoi's mass. To account for the vehicles' nonholonomic nature and velocity constraints, the proposed approach is extended using the minimum curvature-based distance when determining the agent's area of dominance. A set of simulations involving teams of three and five agents showcases the effectiveness of the proposed control approach.

关键词： Surveys Anisotropic control systems Sensor systems Sensors Multi-agent systems Load modeling Autonomous robots

来源：评论

学校读者我要写书评

暂无评论

Decentralized Task Assignment for Networked Unmanned Aerial Vehicles

Decentralized Task Assignment for Networked Unmanned Aerial ...

引用

IEEE Southeastcon

作者： Hava Szarafinski Kiriakos Kiriakidis Department of Weapons Robotics and Control Engineering U.S. Naval Academy Annapolis MD USA

ISBN: (数字)9798331504847

ISBN: (纸本)9798331504854

Competing with the technical advancements of near-peer adversaries, the Navy is shifting towards increased use of Unmanned Aerial Vehicles (UAVs) to accomplish missions. Formations of networked UAVs are susceptible to single-point failure when assigned tasks centrally, leading to the need for decentralization. This paper investigates the optimal allocation of tasks among networked UAVs by means of distributed algorithms to avoid the single-point vulnerability. Furthermore, it addresses the issue of incomplete information about assignment costs, the result of limited direct connections within the formation. The research synthesizes the Hungarian and consensus classes of algorithms to achieve decentralized task allocation subjected to unknown assignment costs and communications being restricted to immediate neighbors. Within the scope of this paper, the number of tasks and agents are equal and agents know their own costs. Assignment costs of other agent-task pairs are not globally known. The success of the proposed approach is established by demonstrating that the synthesized algorithm reaches the same task assignment compared to centralized optimization using the global cost matrix.

关键词： Costs Runtime Network topology Microcontrollers Autonomous aerial vehicles Robustness Topology Timing Resource management Optimization

来源：评论

学校读者我要写书评

暂无评论

Sim2Real Learning With Domain Randomization for Autonomous Guidewire Navigation in Robotic-Assisted Endovascular Procedures

引用

IEEE Transactions on Automation Science and engineering 2025年 22卷 13842-13854页

作者： Yao, Tianliang Wang, Haoyu Lu, Bo Ge, Jiajia Pei, Zhiqiang Kowarschik, Markus Sun, Lining Seneviratne, Lakmal Qi, Peng Tongji University College of Electronics and Information Engineering Shanghai Institute of Intelligent Science and Technology Department of Control Science and Engineering Shanghai200092 China Shanghai University of Engineering Science School of Electronic and Electrical Engineering Shanghai201620 China Soochow University Robotics and Microsystems Center School of Mechanical and Electrical Engineering Jiangsu Suzhou215131 China Advanced Therapies Siemens Healthineers Ltd. Shanghai201318 China University of Shanghai for Science and Technology Oriental Pan-Vascular Devices Innovation College Shanghai200093 China Bavaria Forchheim91301 Germany Khalifa University KUCARS Department of Mechanical and Nuclear Engineering Abu Dhabi United Arab Emirates Tongji University College of Electronics and Information Engineering Shanghai Institute of Intelligent Science and Technology State Key Laboratory of Cardiovascular Diseases and Medical Innovation Center Shanghai East Hospital Department of Control Science and Engineering School of Medicine Shanghai200092 China

Over the past decade, significant advancements have been made in the research and industrialization of robotic systems for endovascular procedures, yet their clinical application remains relatively limited. Physicians commonly report that these robots lack certain intelligent assistive capabilities during procedures. There has been increasing interest and attempts to apply learning-centered algorithms to the training and enhancement of surgical robot skills. This paper proposes an autonomous navigation algorithm for interventional guidewires that is initially trained solely in a virtual simulation environment and subsequently deployed to a real-world robot. Experimental results demonstrate the feasibility of this approach for real-world applications. The proposed approach can help physicians reduce the learning curve for guidewire manipulation and elevate the robot to a higher level of autonomous operation, thereby breaking through the current bottleneck in the level of intelligence for clinical applications of interventional robots. It also holds promise for bringing intelligent transformation to future interventional procedures. © 2004-2012 IEEE.

关键词： Adversarial machine learning

来源：评论

学校读者我要写书评

暂无评论

Retinal disorder diagnosis based on hybrid deep learning models

引用

Multimedia Tools and Applications 2025年 1-25页

作者： Sedik, Ahmed El-Shafai, Walid El-Hag, Noha A. El-Banby, Ghada M. Abd El-Samie, Fathi E. Smart Systems Engineering Laboratory College of Engineering Prince Sultan University Riyadh11586 Saudi Arabia Department of the Robotics and Intelligent Machines Faculty of Artificial Intelligence Kafrelsheikh University Kafr el-Sheikh Egypt Computer Science Department Prince Sultan University Riyadh11586 Saudi Arabia Department of Electronics and Electrical Communications Engineering Faculty of Electronic Engineering Menoufia University Menouf32952 Egypt The Higher Institute of Commercial Sciences Algarbia Al Mahalla Al Kubra31951 Egypt Department of Industrial Electronics and Control Engineering Faculty of Electronic Engineering Menoufia University Menouf32952 Egypt Department of Information Technology College of Computer and Information Sciences Princess Nourah Bint Abdulrahman University P.O. Box 84428 Riyadh11671 Saudi Arabia

Anomaly detection from medical images is badly needed for automated diagnosis. For example, medical images obtained with several modalities, such as magnetic resonance (MR) and confocal microscopy, need to be classified for anomaly detection. The MR images may contain tumors that need to be detected with anomaly detection techniques. In MR imaging, brain tumors are represented as areas composed of abnormal cells that need to be detected. On the other hand, confocal microscopy can be used for retinal imaging. Hence, there is a need to apply classification techniques for diagnosing diseases of the eye, such as Diabetic Retinopathy (DR), which leads to exudates in different regions of the retina. The detection of exudates is necessary for the diagnosis of DR. In this paper, two models are presented for the classification of MR and retinal images. Much research has been done in this area, but it did not achieve high accuracy of classification in an acceptable time. In addition, the previous models did not consider both types of image modalities considered in this paper. The suggested models are Convolutional Neural Network (CNN) and Convolutional Long Short-Term Memory (Conv-LSTM). They are used to classify normal and abnormal cases of both brain tumor and DR diseases. The CNN model is composed of six convolutional layers and six max-pooling layers. A soft-max activation function takes the classification decision. The proposed Conv-LSTM model is composed of three stages: pre-processing, feature extraction, and classification. The deep learning model performance is better than those of traditional networks. The CNN gives an accuracy of 94.26% for DR classification and 95.8% for brain tumor classification, while the Conv-LSTM gives the maximum accuracy of 96.41% for DR classification and 97.4% for brain tumor classification. © The Author(s), under exclusive licence to Springer Science+Business Media, LLC, part of Springer Nature 2025.

关键词： Confocal microscopy

来源：评论

学校读者我要写书评

暂无评论

Global Navigation Satellite System (GNSS) Emulator for Test and Evaluation of Flight controller Performance

Global Navigation Satellite System (GNSS) Emulator for Test ...

引用

International Conference on Unmanned Aircraft Systems (ICUAS)

作者： Matthew McClelland Zachary Cohen Michael D. M. Kutzer Levi DeVries Robotics & Control Engineering Department United States Naval Academy Annapolis Maryland United States of America

ISBN: (数字)9798331513283

ISBN: (纸本)9798331513290

Global navigation satellite system (GNSS) receivers have become ubiquitous geopositioning sensors in unmanned aerial, ground, and surface systems (UxS). GNSS require line of sight communication with orbiting satellites and the resulting measurement's precision and accuracy can be greatly affected by satellite geometry, atmospheric conditions, and obstructions from buildings or foliage, among other effects. These uncertainties can be difficult to manipulate for test and evaluation of a small-scale UAS control system's robustness to GNSS uncertainty. This work presents the implementation of a GNSS emulator with the same interface design as a GNSS receiver. Using a Raspberry Pi connected wirelessly to a local positioning source, we provide a plug-and-play alternative to a standard commercial-off-the-shelf (COTS) GNSS unit that communicates using the DroneCAN protocol. This system allows the user to simulate GNSS measurements and GNSS performance changes by generating synthetic measurements in a controlled laboratory setting. Data collected from outdoor flights in four different environments is used to characterize baseline GNSS message parameter values, which quantify the fix quality in different geographic locations. This information is used to generate synthetic GNSS measurements fed to a Cube autopilot running ArduCopter flight control software in a hardware in the loop simulation. Results show the GNSS emulator can send DroneCAN GNSS messages providing position and fix quality information to the flight controller. These results illustrate how the plug-and-play GNSS emulator can enable test and evaluation of flight controller robustness to uncertainties, signal dropout, and other conditions affecting GNSS measurements in a controlled laboratory environment.

关键词： Global navigation satellite system Uncertainty Satellites Atmospheric measurements Receivers Extraterrestrial measurements Robustness Software measurement Standards Aerospace control

来源：评论

学校读者我要写书评

暂无评论

Meaningful human control of partially automated driving systems: Insights from interviews with Tesla users

引用

Transportation Research Part F: Traffic Psychology and Behaviour 2025年 113卷 213-236页

作者： Suryana, Lucas Elbert Nordhoff, Sina Calvert, Simeon Zgonnikov, Arkady van Arem, Bart Department of Transport and Planning Faculty of Civil Engineering and Geosciences Delft University of Technology Delft Netherlands Electric Vehicle Research Center Institute of Transportation Studies University of California Davis United States Department of Cognitive Robotics Faculty of Mechanical Engineering Delft University of Technology Delft Netherlands Centre for Meaningful Human Control Delft University of Technology Delft Netherlands

Partially automated driving systems are designed to perform specific driving tasks—such as steering, accelerating, and braking—while still requiring human drivers to monitor the environment and intervene when necessary. This shift of driving responsibilities from human drivers to automated systems raises concerns about accountability, particularly in scenarios involving unexpected events. To address these concerns, the concept of meaningful human control (MHC) has been proposed. MHC emphasises the importance of humans retaining oversight and responsibility for decisions made by automated systems. Despite extensive theoretical discussion of MHC in driving automation, there is limited empirical research on how real-world partially automated systems align with MHC principles. This study offers two main contributions: (1) an empirical evaluation of MHC in partially automated driving, based on 103 semi-structured interviews with users of Tesla's Autopilot and Full Self-Driving (FSD) Beta systems;and (2) a methodological framework for assessing MHC through qualitative interview data. We operationalise the previously proposed tracking and tracing conditions of MHC using a set of evaluation criteria to determine whether these systems support meaningful human control in practice. Our findings indicate that several factors influence the degree to which MHC is achieved. Failures in tracking—where drivers' expectations regarding system safety are not adequately met—arise from technological limitations, susceptibility to environmental conditions (e.g., adverse weather or inadequate infrastructure), and discrepancies between technical performance and user satisfaction. Tracing performance—the ability to clearly assign responsibility—is affected by inconsistent adherence to safety protocols, varying levels of driver confidence, and the specific driving mode in use (e.g., Autopilot versus FSD Beta). These findings contribute to ongoing efforts to design partially automated driving

关键词： Human engineering

来源：评论

学校读者我要写书评

暂无评论

FlashMix: Fast Map-Free LiDAR Localization via Feature Mixing and Contrastive-Constrained Accelerated Training

FlashMix: Fast Map-Free LiDAR Localization via Feature Mixin...

引用

IEEE Workshop on Applications of Computer Vision (WACV)

作者： Raktim Gautam Goswami Naman Patel Prashanth Krishnamurthy Farshad Khorrami Department of Electrical and Computer Engineering Control/Robotics Research Laboratory (CRRL) NYU Tandon School of Engineering Brooklyn NY USA

ISBN: (数字)9798331510831

ISBN: (纸本)9798331510848

Map-free LiDAR localization systems accurately localize within known environments by predicting sensor position and orientation directly from raw point clouds, eliminating the need for large maps and descriptors. However, their long training times hinder rapid adaptation to new environments. To address this, we propose FlashMix, which uses a frozen, scene-agnostic backbone to extract local point descriptors, aggregated with an MLP mixer to predict sensor pose. A buffer of local descriptors is used to accelerate training by orders of magnitude, combined with metric learning or contrastive loss regularization of aggregated descriptors to improve performance and convergence. We evaluate FlashMix on various LiDAR localization benchmarks, examining different regularizations and aggregators, and demonstrating its effectiveness for rapid and accurate LiDAR localization in real-world scenarios. The code is available at https://***/raktimgg/FlashMix.

关键词： Location awareness Training Measurement Point cloud compression Laser radar Accuracy Pose estimation Benchmark testing Mixers Convergence

来源：评论

学校读者我要写书评

暂无评论

LiDAR Extrinsic Calibration Without Accessing 3D Point Cloud Data: A New Approach for LiDAR-Vehicle Alignment Inspection

引用

IEEE Sensors Journal 2025年

作者： Song, Hyeong-Seok Park, Jae-Eun Ahn, Chang-Min Kim, Young-Keun Korea Institute of Robotics and Technology Convergence Korea Republic of Handong Global University Department of Mechanical and Control Engineering Pohang Korea Republic of Handong Global University School of Mechanical and Control Engineering Korea Republic of

LiDAR-vehicle external calibration involves estimating the relative pose of the LiDAR body from the ego-vehicle coordinate frame, which is required after the LiDAR is assembled on the vehicle. Current methods necessitate direct wiring to the LiDAR controller unit to access 3D point cloud data. However, situations can limit a third party’s access to the LiDAR control unit for security and other practical reasons. Can the LiDAR-vehicle pose be estimated under such circumstances? Therefore, this study addresses this issue by proposing a new approach to the LiDAR extrinsic calibration method that does not require accessing 3D point cloud data. The proposed solution uses a designed camera-target board system to capture images of LiDAR beams projected on a cylindrical target. By utilizing geometrical modeling equations, it can estimate the relative 6D pose of the LiDAR body on the reference frame. Experiments demonstrated accuracy within 0.05 degrees and 0.7 mm. This one-shot LiDAR-vehicle alignment inspection system when accessing 3D point cloud data is restricted for various practical applications. © 2001-2012 IEEE.

关键词： Cylindrical target LiDAR alignment LiDAR external calibration LiDAR pose estimation NIR laser one-shot calibration point cloud data

来源：评论

学校读者我要写书评

暂无评论

Improving Collision-Free Success Rate For Object Goal Visual Navigation Via Two-Stage Training With Collision Prediction

arXiv

引用

arXiv 2025年

作者： Lian, Shiwei Zhang, Feitian Robotics and Control Laboratory Department of Advanced Manufacturing and Robotics College of Engineering State Key Laboratory of Turbulence and Complex Systems Peking University Beijing100871 China

The object goal visual navigation is the task of navigating to a specific target object using egocentric visual observations. Recent end-to-end navigation models based on deep reinforcement learning have achieved remarkable performance in finding and reaching target objects. However, the collision problem of these models during navigation remains unresolved, since the collision is typically neglected when evaluating the success. Although incorporating a negative reward for collision during training appears straightforward, it results in a more conservative policy, thereby limiting the agent’s ability to reach targets. In addition, many of these models utilize only RGB observations, further increasing the difficulty of collision avoidance without depth information. To address these limitations, a new concept—collision-free success is introduced to evaluate the ability of navigation models to find a collision-free path towards the target object. A two-stage training method with collision prediction is proposed to improve the collision-free success rate of the existing navigation models using RGB observations. In the first training stage, the collision prediction module supervises the agent’s collision states during exploration to learn to predict the possible collision. In the second stage, leveraging the trained collision prediction, the agent learns to navigate to the target without collision. The experimental results in the AI2-THOR environment demonstrate that the proposed method greatly improves the collision-free success rate of different navigation models and outperforms other comparable collision-avoidance methods. Copyright © 2025, The Authors. All rights reserved.

关键词： Deep reinforcement 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：