检索结果-内蒙古大学图书馆

Density-clustering Aggregation for Personalized Federated Learning With AI-Enabled Aerial and Edge Computing in UAVs

引用

IEEE INTERNET OF THINGS JOURNAL 2025年第9期12卷 11220-11232页

作者： Chien, Wei-Che Lin, Chih-Hsun Zhu, Tianli Dai, Cheng Garg, Sahil Mukherjee, Amrit Natl Dong Hwa Univ Dept Comp Sci & Informat Engn Hualien 974301 Taiwan Sichuan Univ Coll Comp Sci Chengdu 610017 Sichuan Peoples R China Ecole Polytech Montreal Elect Engn Dept Montreal PQ H3C 1K3 Canada Canadian Univ Sch Engn Appl Sci & Technol Dept Comp Engn & Computat Sci CECS Dubai U Arab Emirates Chitkara Univ Inst Engn & Technol Ctr Res Impact & Outcome Rajpura 140401 India Univ South Bohemia Fac Sci Dept Comp Sci Ceske Budejovice 37005 Czech Republic

This research introduces the density-clustering-based aggregation for personalized federated learning (DCPFL) algorithm, which utilizes DBSCAN clustering to enhance model accuracy in AI-enabled aerial and edge computing contexts, particularly for uncrewed aerial vehicles (UAVs). The DCPFL framework promotes model sharing among clients, fostering the development of personalized and optimized models. DBSCAN is beneficial in automatically determining cluster numbers using EPS neighborhoods and MinPts, with parameter optimization achieved through cross-experimental analysis. We further refined the model exchange mechanism by integrating a moving average prediction model to optimize the timing of these exchanges. Tests conducted on three public datasets covering two different machine learning (ML) tasks show that DCPFL surpasses existing methods, offering greater accuracy and enhanced adaptability in varied data environments. Implementing this algorithm in UAV networks leverages AI capabilities in aerial and edge computing to efficiently balance personalized modeling requirements with high performance, showcasing its potential to push federated learning (FL) forward in complex and dynamic settings.

关键词： Computational modeling Data models Federated learning Adaptation models clustering algorithms Edge computing Autonomous aerial vehicles Training Timing Accuracy Aerial computing deep mutual learning distributed machine learning (ML) nonindependent personalized federated learning (FL)

来源：评论

学校读者我要写书评

暂无评论

Enhancing Short-Term Load Forecasting Through K-Shape clustering and Deep Learning Integration

引用

IEEE ACCESS 2025年 13卷 30817-30832页

作者： Zhang, Wentao Cheng, Meng Xiang, Qianliang Li, Qinmiao Chongqing Univ Chongqing Univ Univ Cincinnati Joint Coop Inst Chongqing 400044 Peoples R China Hitachi Energy China Ltd Power Consulting Beijing 100015 Peoples R China

Short-term load forecasting (STLF) is essential for the efficient operation and management of modern power grids, impacting dispatch and trading strategies in electricity markets. However, accurately forecasting short-term loads remains challenging due to the difficulty in categorizing diverse operational modes and the limited availability of exogenous variables such as temperature and economic indicators. To address these challenges, this study introduces K-NBEATSx, a novel model that integrates clustering and deep learning methodologies. The methodology begins by using K-Shape clustering to categorize electric load data based on shape similarity, effectively distinguishing different operational modes. Subsequently, the Neural Basis Expansion Analysis With Exogenous Variables (NBEATSx) method is applied by incorporating trend and seasonality modules to enhance forecasting accuracy. Case studies using load datasets from 3 different countries demonstrate that the proposed model outperforms traditional deep learning models across various operational scenarios. Additionally, the integration of clustering algorithms is validated as an effective strategy for improving prediction performance. This research offers an effective new methodology for deep-learning-based STLF, contributing to enhancing the reliability and efficiency of power system operation.

关键词： Load modeling clustering algorithms Time series analysis Predictive models Load forecasting Forecasting Shape Market research Deep learning Data models Short-term load forecasting (STLF) K-shape clustering deep learning NBEATSx power system operation

来源：评论

学校读者我要写书评

暂无评论

An Adaptive clustering Routing Protocol for Wireless Sensor Networks Based on a Novel Memetic Algorithm

引用

IEEE SENSORS JOURNAL 2025年第5期25卷 8929-8941页

作者： Zhang, Wenfen Lan, Yulin Lin, Anping Xiao, Min Xiangnan Univ Sch Comp & Artificial Intelligence Chenzhou 423000 Peoples R China Hunan Engn Res Ctr Adv Embedded Comp & Intelligent Chenzhou 423000 Peoples R China

Wireless sensor networks (WSNs), as an important part of the Internet of Things (IoT), have developed rapidly in recent years. WSNs comprise spatially distributed sensor nodes to collect information from the environment. The energy of nodes is usually limited, so the speed of energy consumption determines the lifespan of the network. This article proposed an energy-efficient adaptive clustering routing algorithm named multiparent differential evolution (MPDE) and variable step-size local search swarm-intelligence-based adaptive clustering routing (VSSLS-SIACR) to minimize the energy consumption of WSNs and prolong the network lifespan. Here, SIACR is a SIACR protocol, while MPDE and VSSLS are a novel memetic algorithms (MA) that belong to the swarm intelligence algorithm. The SIACR protocol classifies WSN nodes into three categories: cluster heads (CHs), cluster members, and free nodes, and employs the MPDE and VSSLS to optimize clustering, with the objectives of reducing energy consumption and balancing the residual energy among nodes. In the comparison experiments, the algorithm presented in this study was compared to the Low-Energy Adaptive clustering Hierarchy (LEACH), LEACH-centralized (LEACH-C), and particle swarm optimization (PSO)-SIACR algorithms for two network topologies with three data fusion rates of 0, 0.5, and 1, respectively. The experimental results revealed that MPDE and VSSLS-SIACR outperformed the other algorithms in terms of network lifetime and residual energy. Meanwhile, it is demonstrated that the proposed algorithm is highly adaptable to different topologies and different data fusion rates. The PSO-SIACR algorithm performed similarly to the MPDE and VSSLS-SIACR algorithms, indicating that the SIACR is, furthermore, an effective framework for clustering routing optimization using swarm intelligence algorithms.

关键词： clustering algorithms Wireless sensor networks Routing Energy consumption Data integration Routing protocols Memetics Particle swarm optimization Optimization Network topology clustering energy-efficient routing memetic algorithm (MA) wireless sensor networks (WSNs)

来源：评论

学校读者我要写书评

暂无评论

HMC-SLAM: A Robust VSLAM Based on RGB-D Camera in Dynamic Environment Combined Hierarchical Multidimensional clustering Algorithm

引用

IEEE TRANSACTIONS ON INSTRUMENTATION AND MEASUREMENT 2025年 74卷

作者： Xu, Bowen Zheng, Zexuan Pan, Zihao Yu, Lei Soochow Univ Dept Mech & Elect Engn Suzhou 215131 Peoples R China

Most of the current visual simultaneous localization and mapping (SLAM) systems assume that the working environment is static, but dynamic targets in the actual working environment will reduce the position estimation accuracy. Aiming at the problem that visual SLAM cannot filter dynamic targets in a dynamic environment, this article proposes a visual SLAM (VSLAM) system for a dynamic environment based on hierarchical multidimensional clustering (HMC) of feature points, called HMC-SLAM, which can effectively reduce the localization error in dynamic environments. The system identifies a dynamic detection box by combining semantic information and polar geometric constraints. In the dynamic detection box, feature point clusters are divided by fusing the polar distance and depth information of feature points. During the process of dynamic feature point cluster removal, the strongest dynamic point cluster is first identified, and then the adjacent point clusters in its depth range are removed together. To further refine the rejection, we define optimized weights for the feature points in the detection box to avoid the influence of potential dynamic points. In the experiments, we evaluate the performance of the proposed HMC-SLAM in the TUM and BONN datasets and real scenarios. In the TUM and BONN datasets, the algorithmic accuracy and stability of HMC-SLAM are improved by 80.37% and 87.48% relative to ORBSLAM3, respectively. The real scenario results show that HMC-SLAM maintains good localization accuracy, effectively rejects dynamic targets, and is more robust than the current state-of-the-art dynamic SLAM methods.

关键词： Heuristic algorithms Dynamics Simultaneous localization and mapping Feature extraction YOLO Cameras clustering algorithms Accuracy Optimization Visualization Depth information dynamic environment hierarchical clustering simultaneous localization and mapping (SLAM)

来源：评论

学校读者我要写书评

暂无评论

Energy-Efficient Hybrid Adaptive clustering for Dynamic MANETs

引用

IEEE ACCESS 2025年 13卷 51319-51331页

作者： Yilmaz, Kudret Kara, Resul Katircioglu, Ferzan Duzce Univ Duzce Vocat Sch Dept Comp Technol TR-81620 Duzce Turkiye Duzce Univ Dept Comp Engn TR-81620 Duzce Turkiye Duzce Univ Dept Mechatron Engn TR-81620 Duzce Turkiye

Mobile ad hoc network (MANET) is a wireless, mobile node network in which the nodes move randomly and operate without centralized management. In MANETs, the network structure increases the energy consumption of the nodes, which shortens the network lifetime and affects packet transmission. The process of clustering in MANETs (Mobile Ad-hoc Network) can be achieved through the division of the network into virtual groups, known as clusters. The Cluster Head (CH) of each cluster is in charge of data transmission within the cluster. In this study, a two-stage Hybrid Adaptive clustering Algorithm for Dynamics MANETs (HACADM) is proposed to improve the network performance in MANETs. In the first stage, based on the Weighted clustering Algorithm (WCA) for selecting optimal CHs, criteria such as node degree, neighborhood distance, power of battery and mobility are optimized using the Gravity Search Algorithm (GSA). In the second phase, the clustering is executed by identifying the member nodes and their roles of the selected CHs using the Enhanced Density Based Spatial clustering of Applications with Noise (Enhanced-DBSCAN) algorithm, which is one of the unsupervised learning methods. Moreover, this approach serves to reduce the load on the CHs and enhance the stability of the cluster by selecting gateway nodes for inter-cluster communication. This study represents a significant step towards optimizing energy efficiency and extending network lifetime by enhancing the adaptability of clustering processes in MANETs under dynamic network conditions. The proposed HACADM method has the potential to enhance the performance of MANETs by ensuring a more balanced load distribution compared to existing clustering approaches. The HACADM method was compared with the EE-WCA, E-MAVMMF, TSDR and MORS-ASO methods using critical performance metrics such as remaining energy, end-to-end delay, packet delivery ratio and throughput. For example, experimental results on remaining energy show tha

关键词： Ad hoc networks Mobile computing clustering algorithms Energy efficiency Energy consumption Routing Heuristic algorithms Routing protocols Mobile nodes Particle swarm optimization Adaptive clustering energy efficiency gravitational search algorithm machine learning MANET

来源：评论

学校读者我要写书评

暂无评论

FWCEC: An Enhanced Feature Weighting Method via Causal Effect for clustering

引用

IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 2025年第2期37卷 685-697页

作者： Cao, Fuyuan Jing, Xuechun Yu, Kui Liang, Jiye Shanxi Univ Sch Comp & Informat Technol Key Lab Computat Intelligence & Chinese Informat P Minist Educ Taiyuan 030006 Shanxi Peoples R China Hefei Univ Technol Sch Comp & Informat Hefei 230009 Peoples R China

Feature weighting aims to assign different weights to features based on their importance in machine learning tasks. In clustering tasks, the existing methods learn feature importance based on the clustering results derived from the collaborative contribution of all features, which overlooks the independent effect of each feature. In fact, there are underlying causal relationships between features and the clustering results, and the features with high causal effects are always more crucial for clustering. Therefore, we propose an enhanced Feature Weighting method via Causal Effect for clustering, calculating the causal effect of each feature on the clustering results for obtaining the independent contribution of each feature. Specifically, we start by identifying the causal relationships among the features and utilizing the causal relationships to generate a reasonable treatment group. Next, we compare the changes in the data distribution between the treatment and control groups to determine the causal effect of each feature. Finally, the causal effects of features are used for enhancing the clustering-driven weight learning. Moreover, we present a theory of relative order consistency in causal effect. Experimental results demonstrate that utilizing causal effect in weight learning facilitates efficient convergence and achieves superior accuracy compared to state-of-the-art clustering algorithms.

关键词： clustering algorithms Machine learning algorithms Machine learning Partitioning algorithms Noise measurement Collaboration Linear programming Convergence Technological innovation Reviews Feature weights clustering causal effect feature importance

来源：评论

学校读者我要写书评

暂无评论

-Graph: A Graph Embedding for Interpretable Time Series clustering

引用

IEEE TRANSACTIONS ON KNOWLEDGE AND DATA ENGINEERING 2025年第5期37卷 2680-2694页

作者： Boniol, Paul Tiano, Donato Bonifati, Angela Palpanas, Themis Inria PSL CNRS ENS F-75005 Paris France Univ Modena I-42121 Modena Italy Lyon 1 Univ IUF LIRIS CNRS Villeurbanne F-69622 Villeurbanne France Univ Paris Cite IUF F-75006 Paris France

Time series clustering poses a significant challenge with diverse applications across domains. A prominent drawback of existing solutions lies in their limited interpretability, often confined to presenting users with centroids. In addressing this gap, our work presents k-Graph, an unsupervised method explicitly crafted to augment interpretability in time series clustering. Leveraging a graph representation of time series subsequences, k-Graph constructs multiple graph representations based on different subsequence lengths. This feature accommodates variable-length time series without requiring users to predetermine subsequence lengths. Our experimental results reveal that k-Graph outperforms current state-of-the-art time series clustering algorithms in accuracy, while providing users with meaningful explanations and interpretations of the clustering outcomes.

关键词： Time series analysis Feature extraction clustering algorithms Accuracy Heuristic algorithms clustering methods Training Shape Partitioning algorithms Directed graphs Time series clustering interpretability

来源：评论

学校读者我要写书评

暂无评论

Rolling Dispatch for AAVs Inspection Based on Task Adaptive clustering

引用

IEEE TRANSACTIONS ON AUTOMATION SCIENCE AND ENGINEERING 2025年 22卷 14108-14119页

作者： Ren, Jize Song, Nanfeng Li, Xian Wang, Lei Beihang Univ Sch Automat Sci & Elect Engn Beijing 100191 Peoples R China Qingdao Univ Sch Automat Qingdao 266071 Shandong Peoples R China

The assignment and scheduling of AAV power inspection tasks constitute a representative Mixed-Integer Nonlinear Programming (MINLP) problem. However, due to the complexity of the workflow, manual methods and existing heuristics struggle to balance time costs and solution quality for large-scale tasks. To address this challenge, this paper introduces a rolling window approach for batch processing and proposes a clustering strategy before scheduling. Building on this strategy, we develop a rolling dispatch algorithm utilizing task-adaptive clustering (RDA-TAC), which alleviates problem complexity and enhances search efficiency. Specifically, we present a novel MINLP model designed to minimize total equivalent endurance losses, incorporating multi-level and sequential inspection constraints derived from actual inspection processes. We then batch and segment large-scale tasks, treating task fragments as the smallest units for assignment, and adaptively adjust them to convert the problem into a single-drone path planning task. Through nine scenarios based on real power grid data, RDA-TAC attains optimal solutions for small-scale tasks and outperforms manual and heuristic methods in more complex cases, demonstrating a tenfold increase in solving speed. Note to Practitioners-Autonomous aerial vehicles (AAVs) are the leading method for power line inspection, offering rapid deployment, cost-effectiveness, and improved safety. However, traditional manual assignment and path planning algorithms often struggle with large-scale and complex tasks. To address this, we propose a strategy that clusters tasks before scheduling, reducing complexity and improving efficiency. Empirical validation demonstrates that our method balances solution time and quality effectively. Comparative results are included in the attached digital format.

关键词： Inspection Drones Autonomous aerial vehicles Routing Poles and towers clustering algorithms Power grids Manuals Costs Adaptation models power inspection task assignment mixed integer nonlinear programming model adaptive clustering optimization

来源：评论

学校读者我要写书评

暂无评论

Dual Completion Learning for Incomplete Multi-View clustering

引用

IEEE TRANSACTIONS ON EMERGING TOPICS IN COMPUTATIONAL INTELLIGENCE 2025年第1期9卷 455-467页

作者： Shen, Qiangqiang Zhang, Xuanqi Wang, Shuqin Li, Yuanman Liang, Yongsheng Chen, Yongyong Harbin Inst Technol Shenzhen Sch Elect & Informat Engn Shenzhen 518055 Guangdong Peoples R China Univ Ottawa Sch Elect Engn & Comp Sci Ottawa ON K1N 6N5 Canada Shandong Univ Aeronaut Coll Sci Binzhou 256600 Peoples R China Shenzhen Univ Coll Elect & Informat Engn Guangdong Key Lab Intelligent Informat Proc Shenzhen 518060 Peoples R China Harbin Inst Technol Shenzhen Sch Comp Sci & Technol Shenzhen 518055 Guangdong Peoples R China

Incomplete Multi-View clustering (IMVC) offers a way to analyze incomplete data, facilitating the inference of unobserved and missing data points through completion techniques. However, existing IMVC methods, predominantly depending on either data completion or similarity matrix completion, failed to uncover the inherent geometric structure and potential complementary information between intra- and inter-views, causing incomplete similarity matrices to further tear apart the connections between views. To address this problem, we propose Dual Completion Learning for Incomplete Multi-view clustering (DCIMC), which elaborately designs data completion and similarity tensor completion, and fuses both of them into a unified model to effectively recover the missing samples and similarities. Concretely, in data completion, DCIMC utilizes subspace clustering to recover the missing and unknown instances directly. Meanwhile, in similarity tensor completion, DCIMC introduces the idea of tensor completion to make better use of the high-order complementary information from multi-view data. By fusing the dual completions, missing information and complementary information in each completion are fully explored by each other, reciprocally enhancing one another to boost the accuracy of our clustering algorithm. Experimental results on various datasets show the effectiveness of the proposed DCIMC. Moreover, our DCIMC also achieved superior or comparable performance in an extended comparison with recent deep learning-based multi-view clustering algorithms.

关键词： Tensors clustering algorithms Laplace equations Matrix decomposition Learning systems Matrix converters Kernel Incomplete multi-view clustering subspace clustering data completion tensor completion

来源：评论

学校读者我要写书评

暂无评论

Adaptive clustering Hierarchical PBFT Algorithm for Secure and Efficient Blockchain-Based IoT Edge Computing

引用

IEEE SENSORS JOURNAL 2025年第2期25卷 3252-3266页

作者： Zheng, Jiali Xie, Jincheng Su, Yanting Guangxi Univ Sch Comp Elect & Informat Guangxi Key Lab Multimedia Commun & Network Techno Nanning 530004 Guangxi Peoples R China

With the rapid development of the Internet of Things (IoT), the importance of data transmission security and efficiency has become increasingly significant. Traditional consensus mechanisms, particularly, the practical Byzantine fault-tolerance (PBFT) algorithm, face notable challenges in the context of blockchain-based edge computing due to scalability issues. Specifically, as the number of nodes increases, PBFT experiences high communication complexity, which leads to significant performance degradation. To tackle this challenge, this article proposes a novel consensus algorithm named adaptive clustering hierarchical-PBFT (ACH-PBFT) that employs a hierarchical cluster structure and node fault-tolerance check to enhance the security and efficiency of blockchain-based edge computing in the IoT. ACH-PBFT adaptively selects an appropriate primary node and forms multiple subnode clusters, facilitating a more scalable consensus process. The system reduces the probability of malicious nodes becoming primary nodes through mutual verification between the primary node cluster and the subcluster consensus. Experimental results indicate that the ACH-PBFT algorithm achieves a 19.83% increase in execution speed, a 23.67% improvement in throughput, and an average reduction of 23.05% in communication overhead compared with a similar double-layer PBFT algorithm lacking hierarchical clustering. In the experiments testing against malicious nodes, the ACH-PBFT algorithm exhibited an average increase of 24.73% in the probability of successfully achieving consensus. Overall, this research provides a practical and effective solution for implementing secure and efficient blockchain-based IoT edge computing using a multilayer architecture to achieve flexible deployment and efficient resource utilization in edge computing environments.

关键词： Blockchains Internet of Things clustering algorithms Edge computing Security Scalability Fault tolerant systems Fault tolerance Consensus protocol Heuristic algorithms Blockchain consensus algorithm Internet-of-Things (IoT) edge computing multilayer architecture practical Byzantine fault tolerance (PBFT)

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：