检索结果-内蒙古大学图书馆

Low frequency residential load monitoring via feature fusion and deep learning

ELECTRIC POWER SYSTEMS RESEARCH 2025年 238卷

作者： Ji, Tianyao Chen, Jiawei Zhang, Luliang Lai, Hongfeng Wang, Jian Wu, Qinghua South China Univ Technol Sch Elect Power Engn Guangzhou 510640 Peoples R China

Non-intrusive load monitoring (NILM) is a technique used to disaggregate the total power signal into individual appliance power signals, which plays an important role in smart grid. Recently, deep learning is widely used to deal with the NILM problem. However, current deep learning models are purely data-driven, which do not consider physical mechanisms, making them less effective in extracting useful features. To address these issues, a new approach for feature extraction based on variational mode decomposition (VMD) and a new deep learning model based on variational autoencoder (VAE) are developed in this paper. The proposed feature extraction approach extracts the pulse feature and concatenates it with the original power data to form multiple features, i.e., which achieves feature fusion to improve the performance of deep learning models better than with a single feature. In addition, a feedback variational mode decomposition (FVMD) is proposed to improve the decomposition performance of the original VMD. The channel attention mechanism is introduced to VAE to improve the performance of the model. To verify the accuracy and robustness of the proposed scheme in NILM, it is compared with the state-of-the-art models on the UK-DALE dataset, and the results show that the proposed feature extraction approach can greatly improve the performance of deep learning models and the proposed new deep learning model outperforms some state-of-the-art models in the realm of NILM.

关键词： Non-intrusive load monitoring Feature fusion Feedback variational mode decomposition variational autoencoder Channel attention mechanism

来源：评论

学校读者我要写书评

暂无评论

Developing a CKD prognostic model: Integrating feature extraction and classifiers for early detection

引用

BIOMEDICAL SIGNAL PROCESSING AND CONTROL 2025年 106卷

作者： Meena, K. Hema, K. GITAM Bengaluru Campus Dept Comp Sci & Engn Bengaluru Karnataka India

In result of its increase in prevalence, chronic kidney disease remains one of the emerging health concerns. Studies have shown that the average survival period of an individual without the functioning of either kidney is usually 18 days, after which transplants or dialysis have to be resorted to. Therefore, early detection becomes important. Diagnosis using ML and DL methods are quite efficient, but their decision processes are not transparent. In medical diagnosis, Machine Learning and DL techniques offer considerably high accuracy in disease prediction. Besides that, analyzing the influence of feature selection and extraction becomes very important in order to come up with high-quality and consistent *** efficacy of these strategies relies on proper feature selection, extraction, and classifiers. The objective of the research is to propose a novel framework to identify the optimal feature for early prediction of CKD. It uses exhaustive feature selection to identify pertinent features, processed by autoencoders and variational auto encoders for dimensionality reduction. This research paper identifies the better classifiers by comparing various ML algorithms such as GB, XGBoost,DT, RF, LR, and KNN and the DL classifier such as CNN for early CKD prediction, using advanced feature extraction *** conducted extensive experimental analysis using a standard dataset collected from Kaggle. The stated approach effectiveness is assessed using various measures, including accuracy, precision, recall, F1-score, and the Jaccard coefficient.

关键词： Chronic Kidney Disease Diagnosis Feature Selection Feature extraction Auto encoder variational autoencoder Machine learning Deep learning

来源：评论

学校读者我要写书评

暂无评论

3D shape reconstruction and generation of natural pozzolan particles

引用

POWDER TECHNOLOGY 2025年 451卷

作者： Peng, Bo Thannasi, Prabu Celik, Kemal New York Univ Abu Dhabi Div Engn POB 129188 Abu Dhabi U Arab Emirates

Natural pozzolans are widely used in the construction industry due to their beneficial properties, including enhanced durability, increased long-term concrete strength, and contributions to sustainability by reducing Portland cement usage and carbon emissions. Additionally, they play a role in producing lunar regolith simulants due to their geochemical similarity to lunar regolith. While their physical and chemical characteristics are wellstudied, the impact of particle morphology is significant. Understanding pozzolan particle shape and surface characteristics can optimize their reactivity, workability, and effectiveness in construction materials. Despite its importance, particle morphology is not widely assessed due to the fine scale of the particles. This paper presents a systematic approach to reconstruct and generate realistic pozzolan particles, offering valuable insights into their morphology and enhancing practical applications. Our proposed method, with its potential to improve numerical studies and serve as a foundation for pozzolan-related applications, holds promise for future construction materials and space applications.

关键词： micro-CT scanning Particle morphology Spherical harmonic analysis Principal component analysis variational autoencoder

来源：评论

学校读者我要写书评

暂无评论

Synthetic data generation in motion analysis: A generative deep learning framework

引用

PROCEEDINGS OF THE INSTITUTION OF MECHANICAL ENGINEERS PART H-JOURNAL OF ENGINEERING IN MEDICINE 2025年第2期239卷 202-211页

作者： Perrone, Mattia Mell, Steven P. Martin, John T. Nho, Shane J. Simmons, Scott Malloy, Philip Rush Univ Med Ctr Dept Orthopaed Surg Sect Young Adult Hip SurgDiv Sports MedRush Med 1620 W Harrison St Chicago IL 60612 USA Univ Missouri Dept Math & Comp Sci St Louis MO USA Arcadia Univ Dept Phys Therapy Glenside PA USA

Generative deep learning has emerged as a promising data augmentation technique in recent years. This approach becomes particularly valuable in areas such as motion analysis, where it is challenging to collect substantial amounts of data. The objective of the current study is to introduce a data augmentation strategy that relies on a variational autoencoder to generate synthetic data of kinetic and kinematic variables. The kinematic and kinetic variables consist of hip and knee joint angles and moments, respectively, in both sagittal and frontal plane, and ground reaction forces. Statistical parametric mapping (SPM) did not detect significant differences between real and synthetic data for each of the biomechanical variables considered. To further evaluate the effectiveness of this approach, a long-short term model (LSTM) was trained both only on real data (R) and on the combination of real and synthetic data (R & S);the performance of each of these two trained models was then assessed on real test data unseen during training. The principal findings included achieving comparable results in terms of nRMSE when predicting knee joint moments in the frontal (R&S: 9.86% vs R: 10.72%) and sagittal plane (R&S: 9.21% vs R: 9.75%), and hip joint moments in the frontal (R&S: 16.93% vs R: 16.79%) and sagittal plane (R&S: 13.29% vs R: 14.60%). The main novelty of this study lies in introducing an effective data augmentation approach in motion analysis settings.

关键词： Generative deep learning variational autoencoder motion analysis musculoskeletal modeling

来源：评论

学校读者我要写书评

暂无评论

RECNN: Restoration and extraction of incomplete brillouin gain spectrum based on CNN framework combined with SVAE and attention mechanism

引用

OPTICS COMMUNICATIONS 2025年 576卷

作者： Shu, Han Zheng, Huan Qin, Yali Zhejiang Univ Technol Coll Informat Engn Hangzhou 310023 Peoples R China

In this paper, we propose a novel method under convolutional neural network framework combined with a supervised variational autoencoder and an attention mechanism, named the Restoration and Extraction Convolutional Neural Network (RECNN), for the restoration and extraction of Incomplete Brillouin Gain Spectrum (IBGS). It is important to clarify that the IBGS discussed here does not encompass information around the Brillouin Frequency Shift (BFS). This omission complicates the task of reconstructing the original spectrum or determining the BFS value. The RECNN framework consists of two main components: Restoration Supervised variational autoencoder (RSVAE) with an attention module for IBGS restoration, and Residual Attention Convolutional Neural Network (RACNN) for BFS extraction. Different types of IBGS are discussed in detail. We define the K index to quantify the incompleteness of the IBGS and introduce the R-squared index to measure the restoration performance of RSVAE. Additionally, the Root Mean Square Error (RMSE) and uncertainty are used to evaluate the overall performance of RECNN. Both simulation and experimental results demonstrate that the Rsquared index increases with increasing K and Signal-to-Noise Ratio (SNR), while both RMSE and uncertainty decrease with increasing K and SNR. In comparisons with various other methods including Linear Curve Fitting (LCF) and artificial neural networks, RECNN consistently outperforms them. Specifically, simulation results show that when K is 0.5 and SNR is 4 dB, the R-squared value for RSVAE reaches 0.82, significantly higher than the 0.31 achieved by LCF method. Experimental results indicate that for an SNR of 6.78 dB and K of 0.5, the RMSE and uncertainty of RECNN are 3.21 MHz and 2.68 MHz, respectively, representing reductions of 11.43 MHz and 11.17 MHz compared to LCF. It's noteworthy that the time consumption evaluation indicates that RECNN requires less than 7 ms to restore the complete BGS and obtain the BFS va

关键词： Brillouin optical time-domain analysis Brillouin gain spectrum Brillouin frequency shift variational autoencoder Convolutional neural network

来源：评论

学校读者我要写书评

暂无评论

A Dataset of 3M Single-DOF Planar 4-, 6-, and 8-Bar Linkage Mechanisms With Open and Closed Coupler Curves for Machine Learning-Driven Path Synthesis

引用

JOURNAL OF MECHANICAL DESIGN 2025年第4期147卷 041702页

作者： Nurizada, Anar Dhaipule, Rohit Lyu, Zhijie Purwar, Anurag SUNY Univ Dept Mech Engn Comp Aided Design & Innovat Lab Stony Brook NY 11794 USA SUNY Univ Dept Comp Sci Comp Aided Design & Innovat Lab Stony Brook NY 11794 USA

In recent years, there has been a strong interest in applying machine learning techniques to path synthesis of linkage mechanisms. However, progress has been stymied due to a scarcity of high-quality datasets. In this article, we present a comprehensive dataset comprising nearly three million samples of 4-, 6-, and 8-bar linkage mechanisms with open and closed coupler curves. Current machine learning approaches to path synthesis also lack standardized metrics for evaluating outcomes. To address this gap, we propose six key metrics to quantify results, providing a foundational framework for researchers to compare new models with existing ones. We also present a variational autoencoder-based model in conjunction with a k-nearest neighbor search approach to demonstrate the utility of our dataset. In the end, we provide example mechanisms that generate various curves along with a numerical evaluation of the proposed metrics.

关键词： planar 4-6-and 8-bar linkage path synthesis variational autoencoder neural networks machine learning deep learning dataset generation data-driven design deep generative models artificial intelligence computational kinematics computer-aided engineering generative design linkages mechanism synthesis

来源：评论

学校读者我要写书评

暂无评论

A novel feature adaptive meta-model for efficient remaining useful life prediction of lithium-ion batteries

引用

JOURNAL OF ENERGY STORAGE 2025年 114卷

作者： Rai, Amit Liu, Jay Pukyong Natl Univ Inst Cleaner Prod Technol Pusan South Korea Pukyong Natl Univ Dept Chem Engn Pusan South Korea

Lithium-Ion Batteries (LiBs) are the most widely used energy storage devices due to their high energy density and long cycle life. However, despite their widespread adoption, the stochastic nature of capacity degradation presents operational and safety challenges, diminishing the remaining useful life (RUL) of the batteries. This research introduces a multi-stage, feature-adaptive meta-model designed to optimize the latent vector space at the meta-data stage, enhancing subsequent meta-model learning. The adaptive nature of the meta-feature space minimizes prediction variance, thereby improving model generalization, prediction accuracy, and computational efficiency, achieving 51.34 % and 85.25 % greater accuracy compared to bagging and boosting methods, respectively. Furthermore, a bidirectional long short-term memory (BiLSTM) and variational autoencoder (VAE)based generative model with an optimized latent dimension is developed to effectively capture statistical variations and temporal dependencies within the RUL dataset, addressing data availability challenges. Additionally, a cost-aware maintenance strategy is formulated, employing a quadratic function to assess the economic impact of precise RUL predictions by penalizing both overestimation and underestimation in different case studies. This study aims to deliver an accurate prediction model, a synthetic data generation method, and a cost-effective maintenance strategy for informed decision-making.

关键词： Lithium-ion batteries Remaining useful life Machine learning BiLSTM variational autoencoder Synthetic data generation Maintenance strategy

来源：评论

学校读者我要写书评

暂无评论

Unsupervised anomaly detection with a temporal continuation, confidence-aware VAE-GAN

引用

PATTERN RECOGNITION 2025年 166卷

作者： Xing, Zeyu Mehmood, Owais Smith, William A. P. Univ York Dept Comp Sci York England Omnicom Balfour Beatty York England

We propose an unsupervised approach to anomaly detection in data with a temporal dimension. We adapt the VAE-GAN architecture to learn the proxy task of temporal sequence continuation. Rather than reconstructing the input, our variational decoder decodes to a forecast of the future sequence. In order to separate structural uncertainty (which our model can reconstruct by fitting to observed data) from stochastic uncertainty (which it cannot) we introduce an additional decoder that outputs the pointwise confidence of the prediction, after the optimal latent-variable has been found. We can use this for zero-shot anomaly detection, separating anomalies from stochastic variation that cannot be modelled, without any examples. This is important for domains in which anomalies are so rare that it is not possible or meaningful to train a supervised model. As an example of such a domain, we introduce a new dataset comprising linescan imagery of railway lines which we use to illustrate our methods. We also achieve state-of-the-art performance on the ECG5000 and MIT-BIH time series anomaly detection datasets. We make an implementation of our method available at https://***/YorkXingZeyu/ECG-VAEGAN-Project.

关键词： Time series anomaly detection Unsupervised anomaly detection variational autoencoder VAE-GAN

来源：评论

学校读者我要写书评

暂无评论

TSC-PCAC: Voxel Transformer and Sparse Convolution-Based Point Cloud Attribute Compression for 3D Broadcasting

引用

IEEE TRANSACTIONS ON BROADCASTING 2025年第1期71卷 154-166页

作者： Guo, Zixi Zhang, Yun Zhu, Linwei Wang, Hanli Jiang, Gangyi Sun Yat sen Univ Sch Elect & Commun Engn Shenzhen Campus Shenzhen 518017 Peoples R China Chinese Acad Sci Shenzhen Inst Adv Technol Shenzhen 518055 Peoples R China Tongji Univ Dept Comp Sci & Technol Shanghai 200092 Peoples R China Ningbo Univ Fac Informat & Sci & Engn Ningbo 315211 Peoples R China

Point cloud has been the mainstream representation for advanced 3D applications, such as virtual reality and augmented reality. However, the massive data amounts of point clouds is one of the most challenging issues for transmission and storage. In this paper, we propose an end-to-end voxel Transformer and Sparse Convolution based Point Cloud Attribute Compression (TSC-PCAC) for 3D broadcasting. Firstly, we present a framework of the TSC-PCAC, which includes Transformer and Sparse Convolutional Module (TSCM) based variational autoencoder and channel context module. Secondly, we propose a two-stage TSCM, where the first stage focuses on modeling local dependencies and feature representations of the point clouds, and the second stage captures global features through spatial and channel pooling encompassing larger receptive fields. This module effectively extracts global and local inter-point relevance to reduce informational redundancy. Thirdly, we design a TSCM based channel context module to exploit inter-channel correlations, which improves the predicted probability distribution of quantized latent representations and thus reduces the bitrate. Experimental results indicate that the proposed TSC-PCAC method achieves an average of 38.53%, 21.30%, and 11.19% bitrate reductions on datasets 8iVFB, Owlii, 8iVSLF, Volograms, and MVUB compared to the Sparse-PCAC, NF-PCAC, and G-PCC v23 methods, respectively. The encoding/decoding time costs are reduced 97.68%/98.78% on average compared to the Sparse-PCAC. The source code and the trained TSC-PCAC models are available at https://***/igizuxo/TSC-PCAC.

关键词： Point cloud compression Image coding Convolution Three-dimensional displays Geometry Transforms Transformers voxel transformer sparse convolution variational autoencoder channel context module

来源：评论

学校读者我要写书评

暂无评论

A generalized zero-shot semantic learning model for batch process fault diagnosis

引用

MEASUREMENT SCIENCE AND TECHNOLOGY 2025年第1期36卷

作者： Liu, Kai Zhao, Xiaoqiang Mou, Miao Hui, Yongyong Lanzhou Univ Technol Coll Elect & Informat Engn Lanzhou Peoples R China Gansu Key Lab Adv Control Ind Proc Lanzhou Peoples R China Lanzhou Univ Technol Natl Expt Teaching Ctr Elect & Control Engn Lanzhou Peoples R China

In industrial monitoring, although zero-shot learning successfully solves the problem of diagnosing unseen faults, it is difficult to diagnose both unseen and seen faults. Motivated by this, we propose a generalized zero-shot semantic learning fault diagnosis model for batch processes called joint low-rank manifold distributional semantic embedding and multimodal variational autoencoder (mVAE). Firstly, joint low-rank representation and manifold learning makes the training samples map to the low-rank space, which obtains the global-local features of the samples while reducing the redundancy in the inputs for the training model;secondly, the bias of human-defined semantic attributes is corrected by predicting the attribute error rate;then, fault samples and corrected semantic vectors are embedded into the consistency space, in which the samples are reconstructed using the mVAE to fully integrate the cross-modal information, meanwhile, Barlow matrix is designed to measure the consistency between the fault samples and the attribute vectors, the higher the consistency, the higher the learning efficiency of attribute classifiers;finally, the generalized zero-shot fault diagnosis experiments are designed and conducted on the penicillin fermentation process and the semiconductor etching process to validate the effectiveness, the results show that the proposed model is indeed possible to diagnose target faults without their samples.

关键词： fault diagnosis batch process generalized zero-shot semantic correction variational autoencoder distributed semantic embedding

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

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

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：