Generalization Capability Research Articles

Data Augmentation for Voiceprint Recognition Using Generative Adversarial Networks

Voiceprint recognition systems often face challenges related to limited and diverse datasets, which hinder their performance and generalization capabilities. This study proposes a novel approach that integrates generative adversarial networks (GANs) for data augmentation and convolutional neural networks (CNNs) with mel-frequency cepstral coefficients (MFCCs) for voiceprint classification. Experimental results demonstrate that the proposed methodology improves recognition accuracy by up to 15% in low-resource scenarios. The optimal ratio of real-to-GAN-generated samples was determined to be 3:2, which balanced dataset diversity and model performance. In specific cases, the model achieved an accuracy of 96.6%, showcasing its effectiveness in capturing unique voice characteristics while mitigating overfitting. These results highlight the potential of combining GAN-augmented data and CNN-based classification to enhance voiceprint recognition in diverse and resource-constrained environments.

Transformer model-based multi-scale fine-grained identification and classification of regional traffic states

To address the limitations in precision of conventional traffic state estimation methods, this article introduces a novel approach based on the Transformer model for traffic state identification and classification. Traditional methods commonly categorize traffic states into four or six classes; however, they often fail to accurately capture the nuanced transitions in traffic states before and after the implementation of traffic congestion reduction strategies. Many traffic congestion reduction strategies can alleviate congestion, but they often fail to effectively transition the traffic state from a congested condition to a free-flowing one. To address this issue, we propose a classification framework that divides traffic states into sixteen distinct categories. We design a Transformer model to extract features from traffic data. The k-means algorithm is then applied to these features to group similar traffic states. The resulting clusters are ranked by congestion level using non-dominated sorting, thereby dividing the data into 16 levels, from Level 1 (free-flowing) to Level 16 (congested). Extensive experiments are conducted using a large-scale simulated traffic dataset. The results demonstrate significant advancements in traffic state estimation achieved by our Transformer-based approach. Compared to baseline methods, our model exhibits marked improvements in both clustering quality and generalization capabilities.

Learning to Train and to Explain a Deep Survival Model with Large-Scale Ovarian Cancer Transcriptomic Data

Background/Objectives: Ovarian cancer is a complex disease with poor outcomes that affects women worldwide. The lack of successful therapeutic options for this malignancy has led to the need to identify novel biomarkers for patient stratification. Here, we aim to develop the outcome predictors based on the gene expression data as they may serve to identify categories of patients who are more likely to respond to certain therapies. Methods: We used The Cancer Genome Atlas (TCGA) ovarian cancer transcriptomic data from 372 patients and approximately 16,600 genes to train and evaluate the deep learning survival models. In addition, we collected an in-house validation dataset of 12 patients to assess the performance of the trained survival models for their direct use in clinical practice. Despite deceptive generalization capabilities, we demonstrated how our model can be interpreted to uncover biological processes associated with survival. We calculated the contributions of the input genes to the output of the best trained model and derived the corresponding molecular pathways. Results: These pathways allowed us to stratify the TCGA patients into high-risk and low-risk groups (p-value 0.025). We validated the stratification ability of the identified pathways on the in-house dataset consisting of 12 patients (p-value 0.229) and on the external clinical and molecular dataset consisting of 274 patients (p-value 0.006). Conclusions: The deep learning-based models for survival prediction with RNA-seq data could be used to detect and interpret the gene-sets associated with survival in ovarian cancer patients and open a new avenue for future research.

Presentation Attack Detection using iris periocular visual spectrum images

In this work, we analyse the comparison between using the periocular area instead of the full face area for Presentation Attack Detection (PAD) in the visual spectrum (RGB). The analysis was carried out by evaluating the performance of five Convolutional Neural Networks (CNN) using both facial and periocular iris images for PAD with two different attack instruments. Additionally, we improved the CNN results by integrating the ArcFace loss function instead of the traditional categorical cross-entropy loss, highlighting that the ArcFace function enhances the performance of the models for both regions of interest, facial and iris periocular areas. We conducted Binary and Multiclass comparisons, followed by cross-database validation to assess the generalization capabilities of the trained models. Our study also addresses some of the current challenges in PAD research, such as the limited availability of high-quality face datasets in the desired spectrum (RGB), which impacts the quality of Presentation Attack Instruments (PAI) examples used in training and evaluation. Our goal was to address the challenge of detecting Iris periocular presentation attacks by leveraging the ArcFace function. The results demonstrate the effectiveness of our approach and provide valuable insights for improving PAD systems using periocular areas in the visual spectrum.

SSMBERT: A Space Science Mission Requirement Classification Method Based on BERT

Model-Based Systems Engineering (MBSE) has demonstrated importance in the aerospace field. However, the MBSE modeling process is often tedious and heavily reliant on specialized knowledge and experience; thus, a new modeling method is urgently required to enhance modeling efficiency. This article focuses on the MBSE modeling in space science mission phase 0, during which the mission requirements are collected, and the corresponding dataset is constructed. The dataset is utilized to fine-tune the BERT pre-training model for the classification of requirements pertaining to space science missions. This process supports the subsequent automated creation of the MBSE requirement model, which aims to facilitate scientific objective analysis and enhances the overall efficiency of the space science mission design process. Based on the characteristics of space science missions, this paper categorized the requirements into four categories: scientific objectives, performance, payload, and engineering requirements, and constructed a requirements dataset for space science missions. Then, utilizing this dataset, the BERT model is fine-tuned to obtain a space science mission requirements classification model (SSMBERT). Finally, SSMBERT is compared with other models, including TextCNN, TextRNN, and GPT-2, in the context of the space science mission requirement classification task. The results indicate that SSMBERT performs effectively under Few-Shot conditions, achieving a precision of 95%, which is at least 10% higher than other models, demonstrating superior performance and generalization capabilities.

Damage Localization and Severity Assessment in Composite Structures Using Deep Learning Based on Lamb Waves

In composite structures, the precise identification and localization of damage is necessary to preserve structural integrity in applications across such fields as aeronautical, civil, and mechanical engineering. This study presents a deep learning (DL)-assisted framework for simultaneous damage localization and severity assessment in composite structures using Lamb waves (LWs). Previous studies have often focused on either damage detection or localization in composite structures. In contrast, this study aims to perform damage detection, severity assessment, and localization using independent DL models. Three DL models, namely the artificial neural network (ANN), convolutional neural network (CNN), and gated recurrent unit (GRU), are compared. To assess their damage detection and localization capabilities. Moreover, zero-mean Gaussian noise is introduced as a data augmentation technique to address the variability and noise inherent in LW signals, improving the generalization capability of the DL models. The proposed framework is validated on a composite plate with four piezoelectric transducers, one at each corner, and achieves high accuracy in both damage localization and severity assessment, offering an effective solution for real-time structural health monitoring. This dual-function approach provides a scalable data-driven method to evaluate composite structures, with applications in predictive maintenance and reliability assurance in critical engineering systems.

Low-Resource Chinese Named Entity Recognition via CNN-based Multitask Learning

Named entity recognition (NER) is a fundamental subtask for information extraction that aims to locate and classify named entities in unstructured text into predefined categories. Recently, large-scale language models (LLMs) have achieved SOTA performance on a variety of natural language processing tasks. However, because NER is a sequence labeling task in nature while LLMs is a text-generation model, the performance of LLMs on NER is still significantly below supervised baselines, and NER remains a difficult task. Meanwhile, the word boundary and semantic information of Chinese words are usually quite vague, as words contained in Chinese texts are not separated by spaces. Thus, the NER task still requires supervised learning paradigm and heavily relies on large amounts of labeled data, such as entity type and boundary information. However, the cost of labeling data can be prohibitively large, and the purely supervised approaches usually suffer from poor generalization capability. In this article, we propose a multitask learning-based bidirectional iterated dilated convolution model, BCNN-CWS, for low-resource NER via leveraging word boundary information of Chinese word segmentation (CWS) task. Specifically, to efficiently recognize named entities, an iterated dilated convolutional model with a limited number of layers is implemented. In addition, a bidirectional causal convolution mechanism is presented for contextual information extraction. Results of extensive experiments on public Chinese datasets demonstrate that BCNN-CWS achieves superior performance over state-of-the-art models, and it yields up to about 50% speed improvement over existing methods. It is worth noting that BCNN-CWS can be further improved by combining with a pretrained model. Received: 25 Spetember 2024 | Revised: 4 November 2024 | Accepted: 28 November 2024 Conflicts of Interest The authors declare that they have no conflicts of interest to this work. Data Availability Statement The data that support the findings of this study are openly available in GitLab at https://github.com/jiangfeng13/BCNN-CWS Author Contribution Statement Tao Wu: Conceptualization, Methodology, Writing – original draft, Writing – review & editing, Visualization, Supervision. Xinwen Cao: Resources, Data curation. Feng Jiang: Software, Validation, Formal analysis, Investigation, Writing – original draft. Canyixing Cui: Data curation, Writing -review & editing. Xuehao Li: Resources. Xingping Xian: Supervision, Project administration, Funding acquisition.

Intelligent aerodynamic modelling method for steady/unsteady flow fields of airfoils driven by flow field images based on modified U-Net neural network

An intelligent modelling method driven by flow field images for predicting steady and unsteady flow filed around aerofoils has been developed. Signed Distance Field (SDF) images achieve dimensionality enhancement of aerofoil geometric information, and ‘synthesised images’ achieve dimensionality enhancement of the angle of attack of the aerofoil and Mach number. An intelligent aerodynamic model for steady flow field of aerofoils is constructed based on the U-Net neural network architecture, and further incorporating a long short-term memory (LSTM) module to construct a U-Net-LSTM neural network architecture to extract the temporal features. Typical NACA aerofoils results show that, the prediction error for steady flow is less than 1.98%, while the prediction error for unsteady flow is less than 2.56%. Additionally, the model demonstrates good generalization capability, with a generalization error for steady flow less than 2.45% and a generalization error for unsteady flow less than 3.34%. This research provides a new method for intelligent aerodynamic modelling based on physical representations. Compared to existing methods, this method avoids the need for extracting aerofoil geometry information and eliminates the necessity of predicting the flow field point by point, making it more concise and efficient. Highlights 1. An aerodynamic model was constructed using U-Net to rapidly predict the steady flow field around airfoils. 2. A Long Short-Term Memory (LSTM) module was incorporated to capture temporal information, enabling the rapid prediction of the unsteady flow field around airfoils. To address the problem of ‘dimension loss’ in the modelling datasets, effective data dimensionality enhancement was achieved using SDF images and ‘synthesized images’.

Optimization of Thermal and Pressure Drop Performance in Circular Pin Fin Heat Sinks Using the TOPSIS Method

This study aims to optimize the thermal performance of pin fin heat sinks by minimizing the maximum temperature of the heat source. Using ANSYS ICEPAK, simulations were conducted for various design parameters, including the number of fins, inlet flow rate, and fin thickness, across circular fins in both inline and staggered arrangements. The circular staggered configuration with 36 fins (3 mm thick) and a flow rate of 6 CFM (Cubic Feet per Minute) achieved the lowest temperature of 34.96 °C, outperforming the inline arrangement. The Taguchi method helped strike a balance between heat transfer and pressure drop, revealing that flow rate has a greater influence when varied compared to the number of fins and fin thickness. An optimal configuration was identified with 36 fins and a flow rate of 4 CFM, which was less sensitive to operational variations. Analysis of Variance (ANOVA) revealed that inlet flow rate significantly impacts heat sink performance, while polynomial regression models demonstrated strong generalization capabilities, with Root mean square error (RMSE) of 8.92%. These findings provide reliable predictive tools and practical insights for optimizing heat sink designs in electronics cooling applications. By utilizing the Technique for Order Preference by Similarity to Ideal Solution (TOPSIS) method, the coefficient of relative closeness (Cn*) is plotted as a main effect. Referring to the multi-objective optimization-based TOPSIS method, it is found that the attributes are partly from the inlet flow rate (Q) are 63.4% of the number of fins (Nf) (25.05%).

ENHANCING WELDING QUALITY THROUGH PREDICTIVE MODELLING — INSIGHTS FROM MACHINE LEARNING TECHNIQUES

In this work, the application of various machine learning (ML) algorithms for predicting tensile strength based on welding parameters in AA2014-T6 aluminium alloy joints is studied. Six ML models namely linear regression, AdaBoost regression, random forest regression, support vector regression (SVR), multi-layer perceptron regression and gaussian process regression (GPR) are considered. The comprehensive analysis revealed that SVR exhibited superior generalization capabilities on unseen data, achieving an R² of 0.89 and a low RMSE of 15.64. In contrast, GPR, despite its high training accuracy, showed significant overfitting. This work highlights the potential of ML in optimizing welding parameters and highlights the importance of model selection and tuning to prevent overfitting and ensure reliable predictions.

Prediction of the Dissolved Oxygen Content in Aquaculture Based on the CNN-GRU Hybrid Neural Network

The dissolved oxygen (DO) content is one of the important water quality parameters; it is crucial for assessing water body quality and ensuring the healthy growth of aquatic organisms. To enhance the prediction accuracy of DO in aquaculture, we propose a fused neural network model integrating a convolutional neural network (CNN) and a gated recurrent unit (GRU). This model initially employs a CNN to extract primary features from water quality parameters. Subsequently, the GRU captures temporal information and long-term dependencies, while a temporal attention mechanism (TAM) is introduced to further pinpoint crucial information. By optimizing model parameters through an improved particle swarm optimization (IPSO) algorithm, we develop a comprehensive IPSO-CNN-GRU-TAM prediction model. Experiments conducted using water quality datasets collected from Eagle Mountain Lake demonstrate that our model achieves a root mean square error (RMSE) of 0.0249 and a coefficient of determination (R2) of 0.9682, outperforming other prediction models with high precision. The model exhibits stable performance across fivefold cross-validation and datasets of varying depths, showcasing robust generalization capabilities. In summary, this model allows aquaculturists to precisely regulate the DO content, ensuring fish health and growth while achieving energy conservation and carbon reduction, aligning with the practical demands of modern aquaculture.

A text classification method combining in-domain pre-training and prompt learning for the steel e-commerce industry

Purpose With the development of Web information systems, steel e-commerce platforms have accumulated a large number of quality objection texts. These texts reflect consumer dissatisfaction with the dimensions, appearance and performance of steel products, providing valuable insights for product improvement and consumer decision-making. Currently, mainstream solutions rely on pre-trained models, but their performance on domain-specific data sets and few-shot data sets is not satisfactory. This paper aims to address these challenges by proposing more effective methods for improving model performance on these specialized data sets. Design/methodology/approach This paper presents a method on the basis of in-domain pre-training, bidirectional encoder representation from Transformers (BERT) and prompt learning. Specifically, a domain-specific unsupervised data set is introduced into the BERT model for in-domain pre-training, enabling the model to better understand specific language patterns in the steel e-commerce industry, enhancing the model’s generalization capability; the incorporation of prompt learning into the BERT model enhances attention to sentence context, improving classification performance on few-shot data sets. Findings Through experimental evaluation, this method demonstrates superior performance on the quality objection data set, achieving a Macro-F1 score of 93.32%. Additionally, ablation experiments further validate the significant advantages of in-domain pre-training and prompt learning in enhancing model performance. Originality/value This study clearly demonstrates the value of the new method in improving the classification of quality objection texts for steel products. The findings of this study offer practical insights for product improvement in the steel industry and provide new directions for future research on few-shot learning and domain-specific models, with potential applications in other fields.

A Machine Learning-Based Approach for Predicting Aerodynamic Coefficients Using Deep Neural Networks and CFD Data

Artificial Intelligence (AI) and Machine Learning (ML) are increasingly being adopted across various fields, including aerodynamics, exhibiting impressive results in complex computational processes and improving prediction accuracy. This study introduces a novel method for airfoil performance assessment through the development and training of a deep Artificial Neural Network (ANN), used for predicting aerodynamic coefficients and pressure distributions, leveraging comprehensive data obtained by using a Computational Fluid Dynamics (CFD) solver. First, an automated CFD solver was developed for obtaining the extensive dataset needed for the effective training of the ANN. The automation process consisted in the generation of a geometry and a mesh, along with the successful integration of the open-source SU2 solver for conducting the aerodynamic simulations, chosen for its versatility and straightforward integration. Once various airfoil analyses were performed and a comprehensive dataset was obtained, data was normalized and the model was trained. Throughout the training process, several model configurations were tested, varying different architectures, hyperparameters and layer settings, until the best-performing layout was chosen. After broad testing and validation, the optimal configuration was identified as being the one to demonstrate the lowest error rates and the most accurate predictions on both training and unseen data, highlighting the model’s generalization capabilities. This Machine Learning-based approach, used as a substitute for traditional methods, provides remarkable accuracy and robustness, capturing complex behaviors and significantly reducing the computational costs associated with CFD simulations.

Real-time pavement distress detection based on deep learning and visual sensors

Pavement ageing significantly affects traffic safety and requires timely maintenance. However, accurately identifying pavement conditions from images captured by in-vehicle cameras remains challenging. This paper proposes a lightweight pavement distress detection method, MASL-YOLO. To address multi-scale defects, the method employs multi-scale channel mix convolution to enhance feature fusion, thereby improving the sensitivity to pavement damage. A multi-path aggregation attention mechanism and deformable convolutions are used to enhance adaptive sampling, strengthening feature extraction in the backbone. Integrating a large-kernel separable structure into the spatial pyramid pooling fusion further enhances the network's generalisation capability in complex environments. Experiments show that MASL-YOLO achieves an mAP of 85.5%, a 5.1% improvement over YOLOv8n and operates at 73.7 frames per second, meeting real-time requirements. The model, integrated with the RealSense D455 visual sensor and the GNSS module on the detection vehicle, achieved fast and effective pavement distress detection in field tests.

Refinement of a kinetic adsorption model through Artificial Intelligence

The adsorption process involves capturing contaminants at active sites on adsorbent materials. Its economic efficiency and simplicity distinguish it as a preferred option amidst concerns about anthropogenic pollution. Traditional kinetic models present challenges in fitting nonlinear behaviors of porous materials, prompting the exploration of alternative approaches. Machine learning has a wide range of applications in various fields, including environmental engineering. These models possess generalization capabilities and are used as predictive tools, but they can lead to undesired behaviors if the system is complex and the model fails to capture this complexity. A novel methodology is proposed where the calibration of traditional models is studied using a machine learning model for adsorption kinetics, with hexavalent chromium as the adsorbate and activated carbon as the adsorbent. Furthermore, the technique of adding synthetic data to influence the model's capacity is studied, considering whether it leads to overfitting.Traditional models include pseudo first and second order kinetics, while multilayer perceptron is used as the machine learning model. The models obtained through the proposed methodology exhibit significant performance compared to traditional models. Additionally, an improved interpretability in their behavior compared to using only an artificial intelligence model is observed. Calibration ensures physical interpretation while enhancing generalization. The incorporation of synthetic data into the artificial intelligence model resulted in overfitting, subsequent ensemble methods effectively leveraged this, reducing bias related errors. The impact of synthetic data on calibration has demonstrated favorable outcomes.

DRGAT: Predicting Drug Responses Via Diffusion-Based Graph Attention Network.

Accurately predicting drug response depending on a patient's genomic profile is critical for advancing personalized medicine. Deep learning approaches rise and especially the rise of graph neural networks leveraging large-scale omics datasets have been a key driver of research in this area. However, these biological datasets, which are typically high dimensional but have small sample sizes, present challenges such as overfitting and poor generalization in predictive models. As a complicating matter, gene expression (GE) data must capture complex inter-gene relationships, exacerbating these issues. In this article, we tackle these challenges by introducing a drug response prediction method, called drug response graph attention network (DRGAT), which combines a denoising diffusion implicit model for data augmentation with a recently introduced graph attention network (GAT) with high-order neighbor propagation (HO-GATs) prediction module. Our proposed approach achieved almost 5% improvement in the area under receiver operating characteristic curve compared with state-of-the-art models for the many studied drugs, indicating our method's reasonable generalization capabilities. Moreover, our experiments confirm the potential of diffusion-based generative models, a core component of our method, to mitigate the inherent limitations of omics datasets by effectively augmenting GE data.

DINOV2-FCS: a model for fruit leaf disease classification and severity prediction

IntroductionThe assessment of the severity of fruit disease is crucial for the optimization of fruit production. By quantifying the percentage of leaf disease, an effective approach to determining the severity of the disease is available. However, the current prediction of disease degree by machine learning methods still faces challenges, including suboptimal accuracy and limited generalizability.MethodsIn light of the growing application of large model technology across a range of fields, this study draws upon the DINOV2 visual large vision model backbone network to construct the DINOV2-Fruit Leaf Classification and Segmentation Model (DINOV2-FCS), a model designed for the classification and severity prediction of diverse fruit leaf diseases. DINOV2-FCS employs the DINOv2-B (distilled) backbone feature extraction network to enhance the extraction of features from fruit disease leaf images. In fruit leaf disease classification, for the problem that leaf spots of different diseases have great similarity, we have proposed Class-Patch Feature Fusion Module (C-PFFM), which integrates the local detailed feature information of the spots and the global feature information of the class markers. For the problem that the model ignores the fine spots in the segmentation process, we propose Explicit Feature Fusion Architecture (EFFA) and Alterable Kernel Atrous Spatial Pyramid Pooling (AKASPP), which improve the segmentation effect of the model.ResultsTo verify the accuracy and generalizability of the model, two sets of experiments were conducted. First, the labeled leaf disease dataset of five fruits was randomly divided. The trained model exhibited an accuracy of 99.67% in disease classification, an mIoU of 90.29%, and an accuracy of 95.68% in disease severity classification. In the generalizability experiment, four disease data sets were used for training and one for testing. The mIoU of the trained model reached 83.95%, and the accuracy of disease severity grading was 95.24%.DiscussionThe results demonstrate that the model exhibits superior performance compared to other state-of-the-art models and that the model has strong generalization capabilities. This study provides a new method for leaf disease classification and leaf disease severity prediction for a variety of fruits. Code is available at https://github.com/BaiChunhui2001/DINOV2-FCS.

HHS-RT-DETR: A Method for the Detection of Citrus Greening Disease

Background: Given the severe economic burden that citrus greening disease imposes on fruit farmers and related industries, rapid and accurate disease detection is particularly crucial. This not only effectively curbs the spread of the disease, but also significantly reduces reliance on manual detection within extensive citrus planting areas. Objective: In response to this challenge, and to address the issues posed by resource-constrained platforms and complex backgrounds, this paper designs and proposes a novel method for the recognition and localization of citrus greening disease, named the HHS-RT-DETR model. The goal of this model is to achieve precise detection and localization of the disease while maintaining efficiency. Methods: Based on the RT-DETR-r18 model, the following improvements are made: the HS-FPN (high-level screening-feature pyramid network) is used to improve the feature fusion and feature selection part of the RT-DETR model, and the filtered feature information is merged with the high-level features by filtering out the low-level features, so as to enhance the feature selection ability and multi-level feature fusion ability of the model. In the feature fusion and feature selection sections, the HWD (hybrid wavelet-directional filter banks) downsampling operator is introduced to prevent the loss of effective information in the channel and reduce the computational complexity of the model. Through using the ShapeIoU loss function to enable the model to focus on the shape and scale of the bounding box itself, the prediction of the bounding box of the model will be more accurate. Conclusions and Results: This study has successfully developed an improved HHS-RT-DETR model which exhibits efficiency and accuracy on resource-constrained platforms and offers significant advantages for the automatic detection of citrus greening disease. Experimental results show that the improved model, when compared to the RT-DETR-r18 baseline model, has achieved significant improvements in several key performance metrics: the precision increased by 7.9%, the frame rate increased by 4 frames per second (f/s), the recall rose by 9.9%, and the average accuracy also increased by 7.5%, while the number of model parameters reduced by 0.137×107. Moreover, the improved model has demonstrated outstanding robustness in detecting occluded leaves within complex backgrounds. This provides strong technical support for the early detection and timely control of citrus greening disease. Additionally, the improved model has showcased advanced detection capabilities on the PASCAL VOC dataset. Discussions: Future research plans include expanding the dataset to encompass a broader range of citrus species and different stages of citrus greening disease. In addition, the plans involve incorporating leaf images under various lighting conditions and different weather scenarios to enhance the model’s generalization capabilities, ensuring the accurate localization and identification of citrus greening disease in diverse complex environments. Lastly, the integration of the improved model into an unmanned aerial vehicle (UAV) system is envisioned to enable the real-time, regional-level precise localization of citrus greening disease.

Cable fault diagnosis with generalization capability using incremental learning and deep convolutional neural network

Power cable faults may exhibit dynamic and differential changes attributable to variations in operating times and environmental conditions. However, the existing artificial intelligence (AI)-based fault diagnosis methods are lack of capability to handle this complexity of fault development. They can only diagnose given faulty types and will malfunction when dealing with newly-emerged faults, which accounts for the lack of generalization capability for AI-based methods. Given that, this paper proposes a cable fault diagnosis method with generalization capability utilizing incremental learning and deep convolutional neural network (DCNN). The idea of knowledge distillation is introduced to reserve the diagnosis information of the DCNN for original faults, and meanwhile, the DCNN is also modified by the classification loss of newly-emerged faulty samples, which equips the DCNN with generalization capability. The category features are obtained utilizing the refined DCNN, and sent to the constructed classifier for accurate incremental fault diagnosis. Experiments are conducted with field and simulated data. The obtained results validate the feasibility and effectiveness of the proposed method.

Modelling icing growth on overhead transmission lines: Current advances and future directions

AbstractThe increasing impact of climate change raises concerns regarding the vulnerability of overhead transmission lines to ice disasters. To address this issue, this study reviews icing growth modelling in two categories: physical‐driven models (PDMs) and data‐driven models (DDMs), covering current advances and future directions. First, PDMs are summarised, focusing on the thermodynamic and fluid mechanics mechanisms. Existing PDMs are compared based on principles, analysing their advantages, disadvantages, and challenges faced. Second, the summarisation of DDMs involves four aspects: data preparation, algorithm selection, model training, and model evaluation. In data preparation, techniques such as preprocessing methods are reviewed to handle multisource data. In algorithm selection, various modelling algorithms are compared and analysed, from basic to deep learning approaches. In model training, processes are summarised to enhance practical applicability, including data partitioning, hyperparameter adjustment, generalisation capability, and model interpretability. In model evaluation, the predictive capabilities are analysed, covering both regression and classification tasks. Subsequently, based on the analyses, a comparison of PDMs and DDMs across various aspects is presented. Finally, future directions in icing growth modelling are outlined. The aim is to enhance icing assessment by understanding the underlying mechanism in attempt to reduce vulnerability and ensure reliability against adverse weather conditions.