检索结果-内蒙古大学图书馆

Wind Farm Meteorological Prediction Model Based on Frequency Domain Feature Extraction Fusion Mechanism

IEEE ACCESS 2025年 13卷 57426-57441页

作者： Liao, Yichen Gao, Ziqi Li, Xufeng Wuhan Univ Technol Sch Econ Wuhan 430070 Peoples R China Wuhan Univ Technol Sch Int Educ Wuhan 430070 Peoples R China

Given the increasing significance of wind energy in the worldwide energy composition, the effective administration of wind farms has emerged as a crucial element in guaranteeing the reliability of the energy provision. Precise forecasting of meteorological conditions is essential for optimizing wind farm operations in this scenario. However, the current prediction model has certain limitations, namely the problem of inaccurate prediction for a long period and the problem of insufficient ability to extract frequency domain information. Among them, we found that frequency domain features are the key to affecting the above problems. Therefore, this study aims to enhance the ability of existing models to capture frequency domain feature information by introducing an innovative frequency domain feature extraction fusion mechanism, thereby significantly improving the accuracy of meteorological condition forecasting. Specifically, this study addresses the issue of long-term dependency in the model by incorporating the Frequency Enhanced Attention module. This integration improves the model's ability to reliably forecast wind farm meteorological conditions over extended periods. Furthermore, this study utilizes the LoRA (Low-Rank Adaptation) fine-tuning strategy to improve the model's generalization capacity across various wind power circumstances. In the experimental part, we verify the model's high-precision forecasting ability on coarse-grained seasonal changes. At the same time, we have extensively verified the model in four different regions, where our model outperformed existing mainstream models (Autoformer, Informer, and Transformer) by an average of 28.9%, 22.7%, and 27.0% respectively, demonstrating the broad applicability of our model. The model prediction results are better than the current mainstream models in critical indicators, proving the model's effectiveness. These experiments validate our model's efficacy and offer novel technical resources for forecasti

关键词： Predictive models Wind farms Frequency-domain analysis Data models Transformers computational modeling Weather forecasting Energy management Meteorological forecasting for wind farms energy management frequency domain modeling Transformer

来源：评论

学校读者我要写书评

暂无评论

Enhancing Federated Learning Robustness Using Locally Benignity-Assessable Bayesian Dropout

引用

IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY 2025年 20卷 2464-2479页

作者： Xue, Jingjing Sun, Sheng Liu, Min Li, Qi Xu, Ke Chinese Acad Sci Inst Comp Technol Beijing 100045 Peoples R China Zhongguancun Lab Beijing 100086 Peoples R China Tsinghua Univ Inst Network Sci & Cyberspace Beijing Peoples R China Tsinghua Univ Beijing Natl Res Ctr Informat Sci & Technol BNRist Beijing Peoples R China Tsinghua Univ Dept Comp Sci & Technol Beijing 100190 Peoples R China

Federated Learning (FL) has emerged as a privacy-preserving training paradigm, which enables distributed devices to jointly learn a shared model without raw data sharing. However, the inaccessible client-side data and unverifiable local training leave FL vulnerable to Byzantine attacks. Most defense strategies focus on penalizing malicious clients in server-side aggregations and ignore clients-side weight units poisoning assessment, failing to maintain robustness and convergence in non-IID settings. In this paper, we propose Federated learning with Benignity-assessable Bayesian Dropout and variational Attention (FedBDA) to achieve local robust training based on fine-grained benignity indicators and guarantee global robustness over non-IID data. Specifically, FedBDA integrates variational inference explanation of dropout into local training, where each client individually quantifies the benign degree of weight units to determine a resilient dropping pattern for the local Bayesian model, enabling client-side robust training with Bayesian interpretability. To accommodate variational distributions of local Bayesian models and globally assess their benign potentials, we design a joint attention mechanism based on Jensen-Shannon divergence among local, global, and median distributions for robust weighted aggregation. Theoretical analysis proves the robustness and convergence of FedBDA. We conduct extensive experiments on four benchmark datasets with five typical attacks, and the results demonstrate that FedBDA outperforms status quo approaches in model performance and running efficiency.

关键词： Bayes methods Training Servers Data models Robustness Distributed databases Uplink Convergence computational modeling Recurrent neural networks Federated learning Byzantine attack dropout defense robust aggregation

来源：评论

学校读者我要写书评

暂无评论

Hierarchical Federated Learning in Inland Waterways via Dataset Distillation and Resource Allocation

引用

IEEE TRANSACTIONS ON VEHICULAR TECHNOLOGY 2025年第3期74卷 3695-3707页

作者： Zhao, Jian Li, Baiyi Yang, Tingting Liu, Jiwei Dalian Maritime Univ DMU Nav Coll Dalian 116026 Peoples R China Peng Cheng Lab Dept Network Intelligence Shenzhen 518066 Peoples R China Dalian Maritime Univ Nav Coll Dalian 116026 Peoples R China

The goal of 6G is to achieve ubiquitous connectivity in heterogeneous and large-scale networks, and Hierarchical Federated Learning (HiFL) is an effective approach to assist AI technologies at the network edge. However, the "waist-shaped client-channel-cloud" architecture of maritime communication network greatly limits the implementation of hierarchical federated edge learning. Considering the aforementioned limitations, a hierarchical federated data distillation method based on UAV-assisted MEC systems is proposed, called Data Augmentation Distilled Federated Learning (DADFL). Specifically, DADFL reduce the communication process between clients and edge servers to just one round. Each client extracts their private dataset, sends synthesized data to the server, and collectively trains a global model. Building upon this, we propose a Global Optimal Iterative Search Algorithm (GOISA) considering client heterogeneity and resource constraints. GOISA minimizes overall latency by optimizing client heterogeneity, bandwidth allocation, and UAV transmission rates. Simulation experiments demonstrate that our proposed system achieves performance comparable to or even better than traditional federated learning, while reducing communication overhead by three orders of magnitude, greatly improving communication efficiency.

关键词： Adaptation models Federated learning computational modeling Computer architecture Channel allocation Data models Servers Resource management Maritime communications Load modeling One-shot federated learning mobile edge computing dataset distillation communication reduction bandwidth allocation

来源：评论

学校读者我要写书评

暂无评论

Digital Triads as Next-Generation Mechatronic Systems for Sustainability-A Case Study

引用

IEEE-ASME TRANSACTIONS ON MECHATRONICS 2025年第1期30卷 679-691页

作者： Bi, Zhuming Mikkola, Aki Devpalli, Dilipkummar Luo, Chaomin Purdue Univ Ft Wayne Dept Civil & Mech Engn Ft Wayne IN 46805 USA Lappeenranta Univ Technol Sch Energy Syst Dept Mech Engn Lappeenranta 53850 Finland Altair Tech Ridge Austin TX 78753 USA Mississippi State Univ Dept Elect & Comp Engn Starkville MS 39759 USA

With the rapid development of information technologies, networking, and artificial intelligence, mechatronics has continuously evolved for several generations from traditional mechatronics to digital twins (DT-I), cyber-physical systems (CPSs), Internet of Things (IoT), human CPSs, and to metaverse nowadays. This article examines such a development trend from a new perspective in enhancing the sustainability of products or systems. To fill the technical gaps of existing concepts in supporting sustainable developments of next-generation mechatronic systems, we propose the concepts of digital triads (DT-II) and Internet of DT-II things;these concepts aim to expand the capabilities of mechatronic systems in dealing with changes and uncertainties in dynamic business environments and prolonging systems' lifespans through technological integration. For example, DT-II is innovative in the sense that it extends the capabilities of CPS or DT-I by introducing a life model so that relevant data, models, and methods can be preserved and transferred when a product or system needs reconfiguring to satisfy new requirements. DT-II promotes concurrent engineering in decision-making supports in the digital and physical worlds so that a product or system can be reconfigured to satisfy new requirements optimally. This smoothens the transitions in sustainable manufacturing and prolongs the enterprise's lifespan. To showcase the use and significance of DT-II, a case study of DT-II for sustainable product development is developed to predict the performance of new products for a client company. The developed DT-II is fully verified and validated for use, and the proposed procedure is generalized and can be used to develop DT-II for other products and systems for enhanced sustainability.

关键词： Mechatronics Sustainable development Manufacturing Business computational modeling Internet of Things Uncertainty Artificial intelligence (AI) cyber-physical systems (CPSs) digital triads (DT-II) digital twins (DT-I) Internet of Digital Triad Things (IoDDT) Internet of Things (IoT) sustainability verification and validation (V&V)

来源：评论

学校读者我要写书评

暂无评论

MFF: A Deep Learning Model for Multi-Modal Image Fusion Based on Multiple Filters

引用

IEEE ACCESS 2025年 13卷 38076-38090页

作者： Wang, Yuequn Li, Zhengwei Wang, Jianli Yang, Leqiang Dong, Bo Zhang, Hanfu Liu, Jie Chinese Acad Sci Changchun Inst Opt Fine Mech & Phys CIOMP Changchun 130033 Peoples R China

Multi-modal image fusion mainly refers to the feature fusion of two or more different images taken from the same perspective range to increase the amount of information contained in an image. This study proposes a multi-modal image fusion deep network called the MFF network. Compared with traditional image fusion models, the MFF network decomposes high-frequency features more finely. In contrast to popular transformer networks, the MFF network utilizes multiple filter networks for the corresponding high and low-frequency feature extraction, thereby improving the model training and inference time. First, GaborNet filtering modules were used by the MFF network to extract high-frequency texture features and invertible neural networks (INN) modules are employed for extracting high-frequency edge features. These two sets of features constitute the high-frequency characteristics of an image. The LEF module is utilized as a low-pass filter to acquire the low-frequency characteristics of an image. The method involving low-frequency feature correlation and high-frequency feature non-correlation was used for image training and fusion purposes. By systematically comparing the TNO, MSRS, and RoadScene datasets with other state-of-the-art image fusion models, the experimental results indicate that the MFF model achieves superior performance in visible-infrared image fusion. Furthermore, evaluations on the LLVIP dataset confirm the model's effectiveness in downstream machine vision tasks. Additionally, comparisons using the MRI_CT, MRI_PET, and MRI_SPECT datasets demonstrate that the MFF model exhibits exceptional performance in medical image fusion.

关键词： Feature extraction Image fusion computational modeling Filters Transformers Deep learning Computer architecture Image edge detection Data mining Convolutional neural networks Autoencoder deep learning filter image fusion visible-infrared image fusion

来源：评论

学校读者我要写书评

暂无评论

Machine Unlearning: Taxonomy, Metrics, Applications, Challenges, and Prospects

引用

IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS 2025年 PP卷 PP页

作者： Li, Na Zhou, Chunyi Gao, Yansong Chen, Hui Zhang, Zhi Kuang, Boyu Fu, Anmin Nanjing Univ Sci & Technol Sch Cyber Sci & Engn Nanjing 210094 Peoples R China Xidian Univ State Key Lab Integrated Serv Networks Xian 710071 Peoples R China Zhejiang Univ Coll Comp Sci & Technol Hangzhou 310058 Peoples R China Univ Western Australia Dept Comp Sci & Software Engn Perth WA 6009 Australia Nanjing Univ Sci & Technol Sch Comp Sci & Engn Nanjing 210094 Peoples R China

Personal digital data is a critical asset, and governments worldwide have enforced laws and regulations to protect data privacy. Data users have been endowed with the "right to be forgotten" (RTBF) of their data. In the course of machine learning (ML), the forgotten right requires a model provider to delete user data and its subsequent impact on ML models upon user requests. Machine unlearning (MU) emerges to address this, which has garnered ever-increasing attention from both industry and academia. Specifically, MU allows model providers to eliminate the influence of unlearned data without retraining the model from scratch, ensuring the model behaves as if it never encountered this data. While the area has developed rapidly, there is a lack of comprehensive surveys to capture the latest advancements. Recognizing this shortage, we conduct an extensive exploration to map the landscape of MU including the (fine-grained) taxonomy of unlearning algorithms under centralized and distributed settings, debate on approximate unlearning, verification and evaluation metrics, and challenges and solutions across various applications. We also focus on the motivations, challenges, and specific methods for deploying unlearning in large language models (LLMs), as well as the potential attacks targeting unlearning processes. The survey concludes by outlining potential directions for future research, hoping to serve as a beacon for interested scholars.

关键词： Data models Surveys Measurement Training Electronic mail Data privacy Taxonomy General Data Protection Regulation computational modeling Approximation algorithms federated learning (FL) large language model (LLM) machine learning (ML) machine unlearning (MU)

来源：评论

学校读者我要写书评

暂无评论

Hard Adversarial Example Mining for Improving Robust Fairness

引用

IEEE TRANSACTIONS ON INFORMATION FORENSICS AND SECURITY 2025年 20卷 350-363页

作者： Lin, Chenhao Ji, Xiang Yang, Yulong Li, Qian Zhao, Zhengyu Peng, Zhe Wang, Run Fang, Liming Shen, Chao Xi An Jiao Tong Univ Sch Cyber Sci & Engn Xian 710049 Peoples R China Xi An Jiao Tong Univ Sch Software Engn Xian 710049 Peoples R China Hong Kong Polytech Univ Dept Ind & Syst Engn TU428 Hong Kong Peoples R China Hong Kong Polytech Univ Shenzhen Res Inst Shenzhen 518057 Peoples R China Wuhan Univ Sch Cyber Sci & Engn Wuhan 430072 Peoples R China Nanjing Univ Aeronaut & Astronaut Sch Comp Sci & Technol Nanjing 210016 Peoples R China

Adversarial training (AT) is widely considered the state-of-the-art technique for improving the robustness of deep neural networks (DNNs) against adversarial examples (AEs). Nevertheless, recent studies have revealed that adversarially trained models are prone to unfairness problems. Recent works in this field usually apply class-wise regularization methods to enhance the fairness of AT. However, this paper discovers that these paradigms can be sub-optimal in improving robust fairness. Specifically, we empirically observe that the AEs that are already robust (referred to as "easy AEs" in this paper) are useless and even harmful in improving robust fairness. To this end, we propose the hard adversarial example mining (HAM) technique which concentrates on mining hard AEs while discarding the easy AEs in AT. Specifically, HAM identifies the easy AEs and hard AEs with a fast adversarial attack method. By discarding the easy AEs and reweighting the hard AEs, the robust fairness of the model can be efficiently and effectively improved. Extensive experimental results on four image classification datasets demonstrate the improvement of HAM in robust fairness and training efficiency compared to several state-of-the-art fair adversarial training methods. Our code is available at https://***/yyl-github-1896/HAM.

关键词： Training Robustness Accuracy modeling Optimization computational efficiency computational modeling Sun Perturbation methods Entropy Adversarial training robust fairness hard adversarial example mining convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

A Systematic Literature Review on Risk Assessment in Cloud Computing: Recent Research Advancements

引用

IEEE ACCESS 2025年 13卷 76289-76307页

作者： Drissi, Saadia Chergui, Meriyem Khatar, Zakaria EMSI Pluridisciplinary Lab Res & Innovat LPRI Casablanca 20110 Morocco Hassan II Univ Comp Sci & Smart Syst Lab C3S Casablanca 20360 Morocco Univ Hassan 2 Super Sch Tech Educ Casablanca 20360 Morocco

The era of Industry 4.0 is characterized by a significant shift in information and communication technology, which has facilitated the emergence of new technology-driven investments. Key components such as Cloud Computing, Artificial Intelligence (AI), Big Data, and the Internet of Things (IoT) form the foundation of this industrial revolution. Cloud Computing serves as a service offering organizations a range of opportunities such as flexibility, scalability, and cost-effectiveness. However, it is essential for organizations to guarantee the security of their resources, especially applications, data, and cloud-based networks to use all cloud offered services effectively for both individuals and organizations. Risk assessment (RA) is essential for identifying, analyzing, and evaluating potential risks that could negatively impact cloud resources, yet developing a tailored approach continues to be challenging, because of the complex nature of cloud. The objective of this systematic literature review is to thoroughly examine the latest advancements in cloud computing security, with a primary focus on RA. Our study explores also various cloud RA scenarios where Machine Learning (ML) and Deep Learning (DL) techniques are applied to identify and mitigate potential risks in cloud environments. This study highlights the latest proposed solutions for assessing risks on cloud, thereby contributing to the development of a comprehensive RA while addressing the specific characteristics of the cloud and its complex nature. Additionally, it addresses process for future research work to enhance RA according many criteria.

关键词： Cloud computing computational modeling Security Organizations Systematic literature review Adaptation models Cloud computing security Software as a service Scalability Systematics Cloud Computing risk assessment security machine learning deep learning risks threats criteria

来源：评论

学校读者我要写书评

暂无评论

Advanced Machine Learning Ensembles for Improved Precipitation Forecasting: The Modified Stacking Ensemble Strategy in China

引用

IEEE JOURNAL OF SELECTED TOPICS IN APPLIED EARTH OBSERVATIONS AND REMOTE SENSING 2025年 18卷 4242-4254页

作者： Tang, Tiantian Wu, Yifan Li, Yujie Xu, Lexi Shi, Xinyi Zhao, Haitao Gui, Guan Nanjing Univ Posts & Telecommun Coll Telecommun & Informat Engn Nanjing 210003 Peoples R China Design Inst Water Conservancy & Hydroelect Power Hangzhou 310000 Peoples R China China United Network Commun Corp Res Inst Beijing 100048 Peoples R China Univ Elect Sci & Technol China Quzhou 324003 Peoples R China Nanjing Great Informat Technol Co Ltd AI Res Ctr Nanjing 210000 Peoples R China

Accurate and reliable precipitation forecasting is vital for effective water resource management and disaster mitigation, especially in geographically diverse and climatically complex regions like China. This study proposes an advanced methodology for medium- and long-term hydrological forecasting by integrating multiple machine learning models through a modified stacking ensemble strategy (MSES). We developed and compared five deterministic precipitation forecasting models, including elastic net regression (ENR), support vector regression, random forest, extreme gradient boosting, and light gradient boosting to provide forecasts with lead times ranging from 0 to 5 months at a spatial resolution of 0.5(degrees). The MSES was then evaluated against the traditional Bayesian model averaging (BMA) approach. Our comprehensive evaluation, based on deterministic forecasting metrics such as the anomaly correlation coefficient (ACC), mean squared skill score (MSSS), and Graded Precipitation Score (Pg), demonstrated the MSES outperformed individual models and the BMA method. The MSES achieved ACC scores between 0.6 and 0.9 for lead time (LT) = 0 month, with an average of around 0.8 for LT = 2 months. The MSSS for MSES was above 0.5 in more than half of the evaluations, and the Pg score was consistently above 80, indicating high accuracy in precipitation magnitude prediction. These findings highlight the promise of advanced machine learning strategies like MSES in improving the accuracy and robustness of precipitation forecasts, addressing critical needs in water resource management and disaster mitigation in China.

关键词： Forecasting Precipitation Predictive models Machine learning Accuracy Radio frequency Training Disasters computational modeling Boosting Ensemble methods machine learning multimodel machine learning integration precipitation forecasting

来源：评论

学校读者我要写书评

暂无评论

Integrated Strategy for Optimized Charging and Balancing of Lithium-Ion Battery Packs

引用

IEEE TRANSACTIONS ON TRANSPORTATION ELECTRIFICATION 2025年第1期11卷 4980-4991页

作者： Astudillo, Galo D. Beiranvand, Hamzeh Cecati, Federico Werlich, Christian Wuersig, Andreas Liserre, Marco Fraunhofer Inst Silicon Technol D-25524 Itzehoe SH Germany Univ Kiel Chair Power Elect D-24143 Kiel SH Germany

During fast charging of lithium-ion batteries (LIBs), cell overheating and overvoltage increase safety risks and lead to faster battery deterioration. Moreover, in conventional battery management systems (BMSs), the cell balancing, charging strategy, and thermal regulation are treated separately at the expense of faster cell deterioration. Hence, this article proposes an optimized fast charging and balancing strategy with electrothermal regulation of LIB packs. Therefore, the power dissipation constraints of passive balancing (PB) are introduced in the proposed integrated optimal framework, and cell balancing is achieved by bypassing the extra charging current. The electrothermal model of the cells, along with a battery pack formed by a string of cells, is implemented. Extensive experiments are carried out to identify the coefficients for the lithium-ion cell model, that is, Samsung-INR18650-20R, and the charging current trajectory as well as the balancing signals are generated with model predictive control (MPC). The pack-level simulations and experiments show that the proposed algorithm maintains the electrothermal boundaries throughout the charging process, increasing the safe charge acceptance of the battery pack.

关键词： Batteries Integrated circuit modeling Resistors Regulation computational modeling Aging Thermal resistance Predictive models Power dissipation Fast charging Battery pack electrothermal battery model fast charging nonlinear model predictive control (NMPC)

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：