IEEE Latin America Transactions最新文献_第2页

TFITS: Time Series Imputation via Dual-Perspective Fusion of Temporal and Feature Views TFITS：基于时间和特征视图双视角融合的时间序列插值

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-04 DOI: 10.1109/TLA.2026.11421262

Junfeng Yuan;Kangyan Li;Baofu Wu;Jian Wan;Jilin Zhang;Yuyu Yin;Yan Zeng

{"title":"TFITS: Time Series Imputation via Dual-Perspective Fusion of Temporal and Feature Views","authors":"Junfeng Yuan;Kangyan Li;Baofu Wu;Jian Wan;Jilin Zhang;Yuyu Yin;Yan Zeng","doi":"10.1109/TLA.2026.11421262","DOIUrl":"https://doi.org/10.1109/TLA.2026.11421262","url":null,"abstract":"In the field of multivariate time series analysis, data completeness plays a crucial role in ensuring the accuracy and reliability of downstream tasks. However, in practical applications, factors such as measurement errors and equipment failures often lead to partial missing values in the data. Therefore, this paper introduces a novel model, focusing on how to efficiently handle missing values in multivariate time series and mitigate the potential negative impact of data incompleteness on subsequent tasks. In previous studies, researchers often adopted a single-perspective approach, separately considering the influence of the temporal dimension and the feature dimension, thereby overlooking the potential of dual-perspective fusion. This paper proposes TFITS, an innovative method for predicting missing values in multivariate time series data. TFITS approaches the problem from both the temporal and feature perspectives of time series data, leveraging the UNetFuse module to fuse feature maps generated from these two perspectives, thereby providing a more comprehensive solution to the missing value problem. Through this approach, TFITS can more effectively address missing values in multivariate time series data. Experimental results demonstrate that the TFITS model not only achieves excellent imputation performance but also exhibits superior and stable performance under different missing rate conditions.","PeriodicalId":55024,"journal":{"name":"IEEE Latin America Transactions","volume":"24 3","pages":"249-259"},"PeriodicalIF":1.3,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11421262","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147351323","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

CEEMDAN-based Single-Channel Blind Source Extraction for Bearing Fault Detection and Diagnosis 基于ceemdan的单通道盲源提取用于轴承故障检测与诊断

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-04 DOI: 10.1109/TLA.2026.11421259

Yuan Xie;Zhangchi Wei;Jinshi Yu;Yifei Sun;Zhipeng Chen

{"title":"CEEMDAN-based Single-Channel Blind Source Extraction for Bearing Fault Detection and Diagnosis","authors":"Yuan Xie;Zhangchi Wei;Jinshi Yu;Yifei Sun;Zhipeng Chen","doi":"10.1109/TLA.2026.11421259","DOIUrl":"https://doi.org/10.1109/TLA.2026.11421259","url":null,"abstract":"Bearings are key components for power transmission in industrial applications. However, how to accurately detect bearing status and perform fault diagnosis in intense background noise remains a challenging task. This paper presents a single-channel blind extraction algorithm for bearing fault detection and diagnosis. First, the vibration signal is decomposed into several intrinsic model functions (IMFs) and a residual. Using a kurtosis index technique to filter the decomposed IMF components, thereby building a multi-channel feature set and overcoming the problem of mode mixing. Then, a blind extraction technology is used to extract fault features from the virtual multi-channel mixed signals. Furthermore, the maximum energy amplitude at the fault feature frequency is obtained to identify fault features by using envelope spectrum analysis. Finally, determining the type of fault in the bearing through the magnitude of singular value entropy to achieve fault diagnosis. Experimental results based on the public bearing fault signal datasets verify the feasibility and superiority. It can make the pulse component in the measurement signal more prominent and suppress the noise component entirely, achieving a powerful de-noising ability. The signal-to-noise ratio has been increased by a maximum of53.35%, and the root mean square error has been decreased by a maximum of 10.86%","PeriodicalId":55024,"journal":{"name":"IEEE Latin America Transactions","volume":"24 3","pages":"280-288"},"PeriodicalIF":1.3,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11421259","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147352548","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

On the building of efficient Fourier Convolutional Neural Networks 高效傅里叶卷积神经网络的构建

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-04 DOI: 10.1109/TLA.2026.11421263

Daniel Lima-López;Pilar Gómez Gil

{"title":"On the building of efficient Fourier Convolutional Neural Networks","authors":"Daniel Lima-López;Pilar Gómez Gil","doi":"10.1109/TLA.2026.11421263","DOIUrl":"https://doi.org/10.1109/TLA.2026.11421263","url":null,"abstract":"In computer vision tasks, Convolutional Neural Networks (CNN) have been the cornerstone to implement numerous models with outstanding results. Nevertheless, its high computational cost hinders its full potential when working with low-end devices. To overcome this drawback, various models have been proposed to take advantage of the properties of the convolution theorem to reduce calculations in CNN architectures. However, the use of the Fourier domain incurs in an increment of the number of weights of the kernels because they are defined using complex variable. In this work, we present a strategy to reduce such numbers by defining a reduced representation, which considers a limited amount of components to keep the parameter count comparable to that of a standard CNN. In deed, we introduce a novel layer, called Linear Transform, that learns to identify the most essential components of a frequency representation, building an appropriate reduced representation. Our experiments on image classification using databases MNIST, Fashion-MNIST, CIFAR-10 and CIFAR-100 showed that our proposal contains fewer parameters and floating point operations, in comparison with both conventional and Fourier CNN architectures. Moreover, the Linear Transform layer produced an improvement in the accuracy in several cases. Cases where accuracy decreases are also analyzed.","PeriodicalId":55024,"journal":{"name":"IEEE Latin America Transactions","volume":"24 3","pages":"239-248"},"PeriodicalIF":1.3,"publicationDate":"2026-03-04","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11421263","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147352528","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Word-Line-Aware Garbage Collector for QLC-based NAND Flash Memories 基于qlc的NAND闪存的字行感知垃圾收集器

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-01 Epub Date: 2026-04-20 DOI: 10.1109/TLA.2026.11488415

Cassiano Silva de Campes;Rodrigo Marques de Figueiredo;Sandro José Rigo

{"title":"Word-Line-Aware Garbage Collector for QLC-based NAND Flash Memories","authors":"Cassiano Silva de Campes;Rodrigo Marques de Figueiredo;Sandro José Rigo","doi":"10.1109/TLA.2026.11488415","DOIUrl":"https://doi.org/10.1109/TLA.2026.11488415","url":null,"abstract":"The Garbage Collector (GC) on NAND Flash memories is one of the most expensive operations in modern Solid-State Drives (SSD). However, it is essential to claim more free pages on SSDs. Various researches attempt to reduce the penalty of GC operations at different levels of the Input/Output (IO) path. Nevertheless, within the reviewed related works in Host-side, Open-Channel SSDs, and Device-side areas, we did not observe overlap with our research scope in how the logical pages are scattered in the NAND Word-Lines (WL) at the physical level, and how they impact performance degradation during the Garbage Collection operation. To reduce this bridge gap, we propose a new GC policy: Word-Line Aware Garbage Collector (WLA-GC). The WLA-GC focuses on reducing the Garbage Collector overhead by observing the physical distribution of valid logical pages within the WLs of a NAND Flash block to best select the victim blocks to be erased. Based on these criteria, the GC decisions are made by considering the latency the victim block will take to be fully erased. Analytical modeling shows that the proposed method outperforms the Vanilla GC in all cases, with an average performance of 25.4%. In comparison, the experimental results present a performance improvement of 55% in best-case scenarios.","PeriodicalId":55024,"journal":{"name":"IEEE Latin America Transactions","volume":"24 5","pages":"476-483"},"PeriodicalIF":1.3,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11488415","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147720872","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Concentration and Diversity in Collaboration Networks in Electrical Engineering: Evidence from Brazilian Researchers 电子工程中协作网络的集中和多样性：来自巴西研究人员的证据

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-01 Epub Date: 2026-04-20 DOI: 10.1109/TLA.2026.11488414

Fernando Almeida do Carmo;Gustavo S. Silva;Raimundo C. S. Freire;Antonio Fernando Lavareda Jacob Junior;Fábio Manoel Franca Lobato

{"title":"Concentration and Diversity in Collaboration Networks in Electrical Engineering: Evidence from Brazilian Researchers","authors":"Fernando Almeida do Carmo;Gustavo S. Silva;Raimundo C. S. Freire;Antonio Fernando Lavareda Jacob Junior;Fábio Manoel Franca Lobato","doi":"10.1109/TLA.2026.11488414","DOIUrl":"https://doi.org/10.1109/TLA.2026.11488414","url":null,"abstract":"Scientific collaboration networks shape research dynamics in fields such as Electrical Engineering, where Brazils CNPq productivity grants play a pivotal role. Expanding upon previous analyzes focused on Microelectronics, this study broadens the scope to encompass the major field of Electrical Engineering, covering seven subareas: Electrical Materials; Electrical, Magnetic, and Electronic Measurements; Instrumentation; Electrical, Magnetic, and Electronic Circuits; Power Systems; Industrial Electronics, Systems, and Electronic Controls; and Telecommunications. The analysis considers collaboration networks among 419 CNPq productivity grant holders based on 18,386 publications. The results reveal a hierarchical structure in which top-tier researchers form a highly influential core, while lower-level researchers remain peripheral. The network exhibits a pronounced geographic concentration in Brazils Southeast and South regions, accompanied by notable gender disparities. Over time, the average author influence has increased, and collaboration clusters have become more compact, suggesting a trend toward smaller, more centralized groups. Despite its thematic diversity, the network exhibits structural inequality and limited cross-level integration. Future research could further investigate temporal shifts, inter-institutional collaboration patterns, and the co-evolution of network structures and research focuses.","PeriodicalId":55024,"journal":{"name":"IEEE Latin America Transactions","volume":"24 5","pages":"456-464"},"PeriodicalIF":1.3,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11488414","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147720873","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Integral Action and Reduced-Order Observer-Based Control for a Magnetic Levitation System 磁悬浮系统的积分作用与降阶观测器控制

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-01 Epub Date: 2026-04-20 DOI: 10.1109/TLA.2026.11488418

Cuauhtémoc Guerrero;Marcos Angel Gonzalez-Olvera;Carlos Micxiu Ortiz-Cervantes

{"title":"Integral Action and Reduced-Order Observer-Based Control for a Magnetic Levitation System","authors":"Cuauhtémoc Guerrero;Marcos Angel Gonzalez-Olvera;Carlos Micxiu Ortiz-Cervantes","doi":"10.1109/TLA.2026.11488418","DOIUrl":"https://doi.org/10.1109/TLA.2026.11488418","url":null,"abstract":"This paper presents the experimental implementation and validation of a single-sensor control architecture for a nonlinear magnetic levitation system based on well-established linear control techniques. The main contribution is the practical demonstration that accurate levitation control can be achieved using only position measurement, thereby simplifying the sensing hardware and reducing implementation costs compared to traditional multi-sensor approaches. An integral action is incorporated into the linear quadratic regulator (LQR) framework to ensure zero steady-state error under step-type reference variations. The complete closed-loop system stability is guaranteed by the separation principle between the controller and the observer dynamics. The proposed design was experimentally validated on a Quanser Magnetic Levitation platform, achieving a low steady-state position error and current estimation error. The experimental results on a Quanser platform demonstrate the practical feasibility of the single-sensor architecture, achieving position errors below 0.35 mm and confirming that classical LQR and reduced-order observer techniques can be effectively integrated for cost-effective magnetic levitation control.","PeriodicalId":55024,"journal":{"name":"IEEE Latin America Transactions","volume":"24 5","pages":"529-537"},"PeriodicalIF":1.3,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11488418","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147720865","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Automatic detection of suicidal ideation on social media using a hybrid CNNBiLSTM architecture 基于混合CNNBiLSTM架构的社交媒体自杀意念自动检测

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-01 Epub Date: 2026-04-20 DOI: 10.1109/TLA.2026.11488419

Sadala Nagib Salame Neto;Kaique da Silva Pinto;Silvio Tadeu Teles da Silva;Miércio Cardoso de Alcântara Neto

{"title":"Automatic detection of suicidal ideation on social media using a hybrid CNNBiLSTM architecture","authors":"Sadala Nagib Salame Neto;Kaique da Silva Pinto;Silvio Tadeu Teles da Silva;Miércio Cardoso de Alcântara Neto","doi":"10.1109/TLA.2026.11488419","DOIUrl":"https://doi.org/10.1109/TLA.2026.11488419","url":null,"abstract":"Suicide is a serious public health problem, responsible for more than 700,000 deaths annually worldwide, according to the World Health Organization. Early identification of signs of suicidal ideation can support prevention initiatives and reduce fatal outcomes. This study investigated the application of Natural Language Processing (NLP) and deep learning techniques for the automatic detection of suicidal ideation in social media posts. A hybrid architecture based on Bidirectional Long Short-Term Memory (BiLSTM), combined with convolutional layers and regularization strategies, was proposed. The evaluation was conducted on a Reddit dataset processed in two versions: with and without stopword removal. The results showed that the proposed architecture achieved high performance in both versions, reaching an accuracy of 95.77% and an F1-score of 95.79% on the test set without stopword removal. Despite the higher computational cost, this approach outperformed previous studies reported in the literature. In summary, the proposed model demonstrates potential to support early monitoring strategies for suicidal ideation in digital environments, offering a relevant contribution to public health and suicide prevention.","PeriodicalId":55024,"journal":{"name":"IEEE Latin America Transactions","volume":"24 5","pages":"465-475"},"PeriodicalIF":1.3,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11488419","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147720868","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0

Hardware-Software Data Acquisition System for Measuring the Grip Strength in Children 儿童握力测量的软硬件数据采集系统

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-01 Epub Date: 2026-04-20 DOI: 10.1109/TLA.2026.11488420

Raquel Ávila Rodríguez;Alejandro Martínez Ramírez;Javier Ávila Rodríguez;Rosa Eréndira Fosado Quiroz;Jorge Luis García Ramírez

引用次数: 0

Table of Contents May 2026 目录2026年5月

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-01 Epub Date: 2026-04-20 DOI: 10.1109/TLA.2026.11488413

引用次数: 0

An Enhanced Fire Perception Framework for Firefighting Robots: ECA-BiFPN Boosted YOLO-EB with Multi-Modal Fusion 一种用于消防机器人的增强火灾感知框架：ECA-BiFPN与多模态融合增强的YOLO-EB

IF 1.3 4区工程技术

IEEE Latin America Transactions Pub Date : 2026-03-01 Epub Date: 2026-04-20 DOI: 10.1109/TLA.2026.11488417

Botao Ni;Lei Huang;Ying Xiang;Yan Zhu;Lin Li;Yunfei Zhou;Jingjing Yang;Hao Tang

{"title":"An Enhanced Fire Perception Framework for Firefighting Robots: ECA-BiFPN Boosted YOLO-EB with Multi-Modal Fusion","authors":"Botao Ni;Lei Huang;Ying Xiang;Yan Zhu;Lin Li;Yunfei Zhou;Jingjing Yang;Hao Tang","doi":"10.1109/TLA.2026.11488417","DOIUrl":"https://doi.org/10.1109/TLA.2026.11488417","url":null,"abstract":"To address the challenges posed by complex fire environments to flame perception and situational assessment in firefighting robots, this paper proposes a multimodal perception method integrating flame segmentation, spatial localization, and situational awareness. Based on an enhanced YOLO-EB segmentation network, this approach combines stereo vision and spectral information to achieve precise flame detection and localization. The YOLO-EB network incorporates an Efficient Channel Attention (ECA) mechanism and a Bidirectional Feature Pyramid Network (BiFPN) to enhance its representational capabilities, effectively balancing accuracy and real-time performance. Ablation experiments demonstrate that the synergistic effect of these modules significantly improves model performance: compared to the YOLOv8-seg baseline, the proposed model achieves a flame recall of 0.680 and a mean average precision (mAP) of 0.821. Leveraging high-quality segmentation masks, a stereo-spectral fusion perception framework is constructed, achieving precise flame localization at medium distances with an average error of 23%. Through dynamic fusion of spectral and visible-light features, the system attains an average situational awareness accuracy of 92.45%, significantly outperforming single-modal methods. Experimental results confirm that this integrated approach provides firefighting robots with stable and reliable capabilities for early fire detection, accurate localization, and dynamic assessment, demonstrating strong potential for practical deployment.","PeriodicalId":55024,"journal":{"name":"IEEE Latin America Transactions","volume":"24 5","pages":"433-444"},"PeriodicalIF":1.3,"publicationDate":"2026-03-01","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"https://ieeexplore.ieee.org/stamp/stamp.jsp?tp=&arnumber=11488417","citationCount":null,"resultStr":null,"platform":"Semanticscholar","paperid":"147720866","PeriodicalName":null,"FirstCategoryId":null,"ListUrlMain":null,"RegionNum":4,"RegionCategory":"工程技术","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":"OA","EPubDate":null,"PubModel":null,"JCR":null,"JCRName":null,"Score":null,"Total":0}

引用次数: 0