检索结果-内蒙古大学图书馆

2024 IEEE Energy Conversion Congress and Exposition, ECCE 2024

作者： Samanta, Akash Chetri, Chandan Karnehm, Dominic Neve, Antje Williamson, Sheldon Ontario Tech University Department of Electrical Computer and Software Engineering Oshawa Canada Bundeswehr University Munich Institute of Intelligent Distributed Systems Neubiberg Germany

ISBN: (纸本)9798350376067

The experimental studies presented in this paper reveal that existing thermal management systems (TMS) and temperature-informed charging algorithms exhibit a response time lag of at least 5.3 minutes due to their reliance on surface temperature measurements. The results indicate that changes in the internal thermal state of lithium-ion batteries (LIBs), induced by variations in charging currents, take an average of 2 minutes to manifest on the battery surface, particularly evident in cylindrical cells. Current thermal management systems for automotive battery packs solely rely on surface temperature measurements, neglecting the approximately 5.8°C temperature difference between the core and surface in TMS control. Consequently, changes in the battery's thermal state due to internal heat losses are not promptly detected by surface-mounted temperature sensors. This delayed response time accelerates battery degradation and increases the risk of thermal runaway events. In this study, temperature-informed fast charging algorithms, tested under various ambient conditions for LIBs, along with a comparative analysis, demonstrate that response time can be reduced by at least 2 minutes by considering internal temperature rather than relying solely on surface temperature measurements. Moreover, accounting for the temperature difference between the core and surface facilitates rapid TMS control and health-conscious fast charging, thereby mitigating the risk of thermal runaway events. © 2024 IEEE.

关键词： Surface measurement

来源：评论

学校读者我要写书评

暂无评论

A novel deep learning approach for automated grading of knee osteoarthritis severity

引用

Multimedia Tools and Applications 2024年 1-20页

作者： Kaur, PrabSimran Kohli, Guneet Singh Bedi, Jatin Wasly, Saud Department of Computer Science and Engineering Thapar Institute of Engineering and Technology Punjab Patiala India Electrical and Computer Engineering Department King Abdulaziz University Jeddah22254 Saudi Arabia Center of Excellence in Intelligent Engineering Systems King Abdulaziz University Jeddah22254 Saudi Arabia

Knee Osteoarthritis (OA) is a prevalent musculoskeletal disorder that affects the knee joint that causes pain, stiffness, and reduced mobility. It is also known as "Degenerative Joint Disease" and is caused by the degeneration of cartilage in the knee joint, leading to bone-on-bone contact and further damage. Knee OA is prevalent in the population, affecting around 22% to 39% of people in India, and there is currently no treatment available that can halt the progression of the disease. Therefore, early diagnosis and management of symptoms are essential to reduce its impact on an individual’s quality of life. To address this issue, have introduced a framework that leverages ConvNeXt architecture, a modernization of ResNets (ResNet-50) architecture towards Hierarchical Transformers (Swin Transformers), to provide accurate identification and classification of knee osteoarthritis. The classification of knee osteoarthritis was done using the Kellgren and Lawrence (KL) graded X-ray images. These images of the damaged knees are preprocessed and augmented, creating a scaled, enhanced, and varied version of the features, thus making the data fitter and more significant for classification. The performance estimation of the proposed strategy is conducted on the Osteoarthritis Initiative (OAI), a research project focused on knee osteoarthritis that works in partnership with NIH and other private industries to develop a public domain dataset that can facilitate research and evaluation. It involves training the prepared data using various hyper-tuned versions of ConvNeXt. The different fine-tuned results of the ConvNeXt models on each KL Grade are evaluated against the other state-of-the-art models and vision transformers. The comparative assessment of widely used performance measures shows that the proposed approach outperforms the conventional models by generating the highest score for all the KL grades. Lastly, an approach is employed to statistically confirm the validity of t

关键词： Joints (anatomy)

来源：评论

学校读者我要写书评

暂无评论

Quantum Calculus-Based Two-Dimensional Least Mean Square Algorithm 58

Quantum Calculus-Based Two-Dimensional Least Mean Square Alg...

引用

58th Asilomar Conference on Signals, systems and Computers, ACSSC 2024

作者： Moinuddin, Muhammad Zerguine, Azzedine Al-Saggaf, Ubaid M. Arif, Muhammad King Abdulaziz University Electrical and Computer Engineering Department The Centre of Excellence in Intelligent Engineering Systems Jeddah Saudi Arabia Kfupm Electrical Engineering Department The Interdisciplinary Research Center for Communication Systems and Sensing Dhahran31261 Saudi Arabia Institute of Networked and Embedded Systems University of Klagenfurt Klagenfurt Austria The Ubiquitous Sensing Systems Lab University of Klagenfurt-Silicon Austria Labs Klagenfurt Austria

ISBN: (纸本)9798350354058

Recent work of the one-dimensional q-LMS algorithm give insight that the Quantum Calculus based gradient called the q-gradient is capable of improving the convergence behavior in contrast to the standard LMS algorithm. The reason for this is the fact that the q-derivative takes larger steps in minimizing a cost function as it evaluates the secant in contrast to the tangent of the cost function. Motivated by this, we propose a two-dimensional version of the q-LMS algorithm. More precisely, we employ Quantum Calculus based q-gradient in steepest descent optimization of mean-square-error (MSE) cost function in 2D adaptive filtering scenario to obtain two-dimensional q-LMS (2D q-LMS) algorithm. A thorough analytical investigation of the proposed 2D q-LMS algorithm in both mean and mean-square sense is provided. Both the transient and steady-state convergence behaviours are examined. Consequently, excess MSE (EMSE) learning curve and its steady-state expression are evaluated in closed form. Simulation results are presented for the application of noise cancelation in images to show the superiority of the proposed 2D q-LMS algorithm. © 2024 IEEE.

关键词： Cost functions

来源：评论

学校读者我要写书评

暂无评论

Photovoltaic Energy Usage for Agricartulal Building: A case Study 23

Photovoltaic Energy Usage for Agricartulal Building: A case ...

引用

23rd International Symposium on electrical Apparatus and Technologies, SIELA 2024

作者： Radeva, Tatyana Petkova, Nikolina Institute of Information and Communication Technologies Bulgarian Academy of Sciences Intelligent Systems Department Sofia Bulgaria Technical University of Sofia Department of Theoretical Electrical Engineering Sofia Bulgaria

ISBN: (纸本)9798350382365

This paper presents a case study of a photovoltaic energy system for an agricultural building. The purpose of the case study is to provide technical measures to improve the energy consumption of the building, according to the technical specification and the current regulations. Agricultural production buildings are by and large used during daytime, where the solar energy will be used only for own consumption. An analysis of the energy consumption for the building is done and a daily assessment shows the optimal requirements for such systems. The obtained outcome can be used in the development self-sustaining power supply systems for different types of buildings, concepts for buildings with zero and almost zero energy consumption. © 2024 IEEE.

关键词： Green buildings

来源：评论

学校读者我要写书评

暂无评论

Active Camera View Planning Based on Deep Reinforcement Learning in Indoor Exploration 42

Active Camera View Planning Based on Deep Reinforcement Lear...

引用

42nd Chinese Control Conference, CCC 2023

作者： Wang, Hao-Yu Meng, Qing-Hao Dai, Xu-Yang Institute of Robotics and Autonomous Systems School of Electrical and Information Engineering Tianjin University Tianjin Key Laboratory of Intelligent Unmanned Swarm Technology and System Tianjin300072 China

ISBN: (纸本)9789887581543

In the last decades, significant progress has been made in the field of active vSLAM (visual simultaneous localization and mapping). However, the majority of active vSLAM methods suffer from the problem of coupling the camera's viewpoint to the robot's motion, which limits the initiative during the exploration of unknown environments. To address this issue, this study proposes a camera view planning method based on deep reinforcement learning (DRL), which makes the camera rotate independently. First, feature encoding is performed on the reconstructed navigation map using a multi-layer convolutional neural network (CNN). Second, the DRL-based camera view planning method is trained through the proximal policy optimization (PPO) algorithm. Last, a significant number of tests is performed on a public dataset of indoor environments. The simulation results show that the proposed camera view planning method improves the exploration coverage by an average of 4% (33.86 m2 for traditional vs. 35.21 m2 for the suggested method). Meanwhile, the kind of semantic information newly observed in each step increases by 21% (2.9 for traditional vs. 3.5 for the proposed method) on average, indicating the improvement in the efficiency of information acquisition. © 2023 Technical Committee on Control Theory, Chinese Association of Automation.

关键词： Reinforcement learning

来源：评论

学校读者我要写书评

暂无评论

Toward better semantic segmentation by retaining spectral information using matched wavelet pooling

引用

Neural Computing and Applications 2025年第10期37卷 7049-7066页

作者： El-Khamy, Said El-Bana, Shimaa Al-Kabbany, Ahmad Elragal, Hassan Department of Electrical Engineering Alexandria University Alexandria21544 Egypt Intelligent Systems Lab Arab Academy for Science Technology and Maritime Transport Alexandria21937 Egypt Department of Research and Development VRapeutic Inc. Cairo11613 Egypt Alexandria Higher Institute of Engineering and Technology Alexandria21311 Egypt

Pooling operations, such as average pooling, strided convolution, and max pooling, have become fundamental components of convolutional neural networks (CNNs) due to their ability to capture local features, expand receptive fields, and reduce computational costs. However, in the context of semantic segmentation, these pooling techniques can lead to the loss of crucial spatial details that are necessary for accurate pixel-level predictions. To tackle this issue, extensive research has focused on refining deep CNN models through architectural adaptations and novel training methods. Recent studies have demonstrated the importance of pooling layers, exemplified by innovations like the introduction of wavelet pooling. In our study, we highlight the value of incorporating our previously proposed matched wavelet pooling (MWP) into CNNs to enhance semantic segmentation pipelines. The core concept of MWP challenges the notion that including all sub-bands generated from wavelet decomposition consistently improves accuracy. Instead, we advocate for selecting specific sub-bands for the pooling process in each image during both training and testing. This approach introduces sub-band selection protocols customized for image-specific pooling, designed specifically for semantic segmentation CNN architectures, with a particular focus on the UNet and SegNet models. Across three widely used datasets, our proposed MWP- based pipeline, featuring the MWP-UNet architecture, consistently outperforms conventional pooling methods. It achieves a significant average improvement in intersection over union (IoU) of over 25% compared to recent literature. Additionally, our MWP-SegNet model outperformed the standard SegNet by 12.5% mIoU, further demonstrating the effectiveness of our matched wavelet pooling approach across different network architectures. © The Author(s) 2025.

关键词： Semantic Segmentation

来源：评论

学校读者我要写书评

暂无评论

Feature Space Augmentation and Old Class Space Preservation for Class Incremental Learning 42

Feature Space Augmentation and Old Class Space Preservation ...

引用

42nd Chinese Control Conference, CCC 2023

作者： Zeng, Ming Li, Siying Liu, Jiazhe Li, Yuanhao Institute of Robotics and Automation Systems School of Electrical and Information Engineering Tianjin Key Laboratory of Intelligent Unmanned Swarm Technology and System Tianjin University Tianjin300072 China

ISBN: (纸本)9789887581543

The ability of continuously learning new knowledge without forgetting old ones is crucial to adapt to an ever-changing world. This scenario becomes more challenging when the previous data are not available. Current class incremental learning (CIL) approaches tend to incorporate new classes with backward compatibility, i.e., maintaining discriminability of old ones. By contrast, we focus on the extensibility and compatibility for future new classes in the early stage. Our proposed method achieves this by augmenting feature space. In detail, a large number of pseudo-new classes are generated via real example mixture and then train the initial spatially augmented model using pseudo-new classes and the base classes. Besides, considering the fact that it is impossible to maintain backward compatibility with only one prototype for each old class, the updated model also needs to preserve old class space in incremental stages. We employ a certain perturbation to the old class prototypes to effectively avoid old class space from being over-squeezed by the samples of new classes. The experiments on CIFAR-100 and ImageNet-Subset (100 classes) demonstrate that our method substantially reduces the overlap of old and new classes, outperforming state-of-the-art various baselines. © 2023 Technical Committee on Control Theory, Chinese Association of Automation.

关键词： Catastrophic forgetting Class incremental learning Feature space augmentation Old class space preservation Pseudo-new classes

来源：评论

学校读者我要写书评

暂无评论

Motion Capture and Gait Analysis of Ants with Leg Injuries 42

Motion Capture and Gait Analysis of Ants with Leg Injuries

引用

42nd Chinese Control Conference, CCC 2023

作者： Zeng, Ming Li, Yuanhao Meng, Chang Zhong, Shutong Wang, Zhijing Zhao, Feng Institute of Robotics and Automation Systems School of Electrical and Information Engineering Tianjin University Tianjin Key Laboratory of Intelligent Unmanned Swarm Technology and System Tianjin300072 China

ISBN: (纸本)9789887581543

Many insects in nature, such as ants, recover quickly after sustaining leg injuries. If this capability could be imparted to multi-legged robots, it would drastically enhance their ability to resist destruction and adapt to the environment. In this paper, the Camponotus Japonicus is used as the research object, and its locomotive patterns are studied under different leg-loss conditions. To facilitate this, a high-speed acquisition system, with a dual-view (bottom view and side view) setup, is designed to analyze the ant's linear gaits. Subsequently, the pivotal gait characteristics of the injured ants are extracted by keyframe annotation technology, including the gait phase diagram, duty factor, angle before support, angle before swing, and leg swing range. Lastly, according to the obtained gait characteristics, the general and specific laws of gait adjustment in the absence of one or two legs are systematically analyzed, which provides the basis for parametric settings and strategy optimization of robotic injured gait self-adjustment. © 2023 Technical Committee on Control Theory, Chinese Association of Automation.

关键词： Gait analysis

来源：评论

学校读者我要写书评

暂无评论

A deep learning-based model for automated STN localization using local field potentials in Parkinson’s disease

引用

Soft Computing 2025年第4期29卷 2343-2362页

作者： Hosny, Mohamed Naeem, Mohamed A. Zhu, Minwei Gao, Wenpeng Elshenhab, Ahmed M. Fu, Yili Department of Electrical Engineering Benha Faculty of Engineering Benha University Benha Egypt State Key Laboratory of Robotics and Systems Harbin Institute of Technology Harbin150080 China Department of Mechatronics and Robotics Engineering Faculty of Engineering South Valley University Qena Egypt First Affiliated Hospital Harbin Medical University Harbin150001 China School of Life Science and Technology Harbin Institute of Technology Harbin150080 China Department of Mathematics Faculty of Science Mansoura University Mansoura35516 Egypt

Accurate identification of the subthalamic nucleus (STN) borders is time-consuming, relying heavily on the neurosurgeon expertise in manually interpreting the electrophysiological signals. Local field potentials (LFPs) have garnered imperative attention due to their strong correlation with the STN. However, existing detection models often face challenges with high computational complexity, hyperparameter optimization and lack of explainability, making them unreliable for clinicians. Therefore, this study introduces an explanatory framework using convolutional neural networks (CNN) for detecting the STN region from LFPs. Continuous wavelet transform is employed to convert LFPs signals into scalogram images, which are then processed by sixteen CNN models. We evaluated our framework by examining the impact of various limiting factors on the classification performance, including model size, learning rate (LR), optimizers and data scaling. Deep features are extracted from the top-performing CNN architectures to capture rich representations of the scalograms. These features are then fused and classified using k-nearest neighbour algorithm. Gradient-weighted class activation mapping is used to explain the decisions made by the proposed model. Our approach achieved an accuracy of 99.61%, outperforming individual CNN models for STN localization. The experimental results revealed that CNN models, embedded with additional hyperparameters and layers, generally outperformed smaller models. Besides, low LR significantly enhanced the performance compared to high LR. Moreover, features extracted from untuned networks produced lower performance than tuned networks. The proposed system could revolutionize deep brain stimulation surgery by increasing efficiency and reducing reliance on clinician expertise for STN detection. © The Author(s), under exclusive licence to Springer-Verlag GmbH Germany, part of Springer Nature 2025.

关键词： Wavelet transforms

来源：评论

学校读者我要写书评

暂无评论

Integrating Machine Learning and Mathematical Optimization for Job Shop Scheduling

引用

IEEE Transactions on Automation Science and engineering 2024年第3期21卷 4829-4850页

作者： Liu, Anbang Luh, Peter B. Sun, Kailai Bragin, Mikhail A. Yan, Bing Department of Automation and BNRist Beijing100084 China University of Connecticut Department of Electrical and Computer Engineering StorrsCT06269 United States National Taiwan University Department of Electrical Engineering Taipei10617 Taiwan Tsinghua University Chaired Professors Group Center for Intelligent and Networked Systems Department of Automation Beijing100084 China Rochester Institute of Technology Department of Electrical and Microelectronic Engineering RochesterNY14623 United States

Job-shop scheduling is an important but difficult combinatorial optimization problem for low-volume and high-variety manufacturing, with solutions required to be obtained quickly at the beginning of each shift. In view of the increasing demand for customized products, problem sizes are growing. A promising direction is to take advantage of Machine Learning (ML). Direct learning to predict solutions for job-shop scheduling, however, suffers from major difficulties when problem scales are large. In this paper, a Deep Neural Network (DNN) is synergistically integrated within the decomposition and coordination framework of Surrogate Lagrangian Relaxation (SLR) to predict good-enough solutions for subproblems. Since a subproblem is associated with a single part, learning difficulties caused by large scales are overcome. Nevertheless, the learning still presents challenges. Because of the high-variety nature of parts, the DNN is desired to be able to generalize to solve all possible parts. To this end, our idea is to establish 'surrogate' part subproblems that are easier to learn, develop a DNN based on Pointer Network to learn to predict their solutions, and calculate the solutions of the original part subproblems based on the predictions. Moreover, a masking mechanism is developed such that all the predictions are feasible. Numerical results demonstrate that good-enough subproblem solutions are predicted in many iterations, and high-quality solutions of the overall problem are obtained in a computationally efficient manner. The performance of the method is further improved through continuous learning. Note to Practitioners - Scheduling is important for the planning and operation of job shops, and high-quality schedules need to be obtained quickly at the beginning of each shift. To take advantage of ML, in this paper, a DNN is integrated within our recent decomposition and coordination approach to learn to predict 'good-enough' solutions to part subproblems. To be able t

关键词： Deep neural networks

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：