多尺度局部特征和Transformer全局学习融合的发动机剩余寿命预测

飞机发动机剩余寿命(Remaining useful life,RUL)的准确预测对确保其安全性和可靠性至关重要.在基于多传感器检测数据预测时,需解决局部特征提取问题以全面捕捉设备在不同时间尺度下的退化趋势,并需解决时间序列中各元素之间长期依赖性的全局学习问题.因此,提出了结合多尺度局部特征增强单元(Multi-sacle local feature enhancement unit,MSLFU_BLOCK)和Transformer编码器的预测模型,称之为MS_Transformer.MSLFU_BLOCK利用堆叠的因果卷积逐层从时间序列数据中提取多尺度局部信息,同时避免了传统卷积计算中固有的未来数据泄漏问题.随后,Transformer编码器通过其自注意机制进一步捕获时间序列数据中的短期和长期依赖关系.通过将多尺度局部特征增强单元与Transformer编码器相结合,提出的MS_Transformer全面捕捉了时间序列数据中的局部和全局模式.在广泛使用的CMAPSS基准数据集上进行的消融和预测实验验证了模型的合理性和有效性.与13个先进预测模型的比较分析表明,MS_Transformer模型在操作条件更复杂的FD002和FD004数据集上的RMSE和Score指标优于其他模型,同时在四个数据集上的平均性能最优.该研究为发动机剩余寿命预测提供了更为可靠的解决方案.

结合CSWin-Transformer和门卷积的壁画图像修复方法

敦煌壁画是珍贵的文化遗产,但现存壁画存在着大量破损现象。针对现有图像修复方法在处理敦煌壁画时面临着计算复杂度高、纹理模糊和特征提取不足等问题,提出了一种结合CSWin-Transformer(cross stripe window-Transformer)和门卷积的壁画图像修复方法。构建由全局层网络和局部层门卷积残差密集网络组成的并行网络,利用条纹窗口增强图像特征提取能力,并通过门卷积残差块提升结构纹理修复的准确性。设计全局-局部特征融合模块来融合全局层和局部层输出的特征图像,以保持修复结果整体的一致性。通过建立共享注意力机制实现全局层和局部层之间的信息交互,同时为了完成破损壁画的修复,采用谱归一化马尔科夫判别模型进行对抗训练。通过对真实破损壁画的修复实验,结果表明,所提方法在主客观指标上均优于所对比的方法。

融合Transformer和交互注意力网络的方面级情感分类模型

现有的大多数研究者使用循环神经网络与注意力机制相结合的方法进行方面级情感分类任务。然而,循环神经网络不能并行计算,并且模型在训练过程中会出现截断的反向传播、梯度消失和梯度爆炸等问题,传统的注意力机制可能会给句子中重要情感词分配较低的注意力权重。针对上述问题,该文提出了一种融合Transformer和交互注意力网络的方面级情感分类模型。首先利用BERT(bidirectional encoder representation from Transformers)预训练模型来构造词嵌入向量,然后使用Transformer编码器对输入的句子进行并行编码,接着使用上下文动态掩码和上下文动态权重机制来关注与特定方面词有重要语义关系的局部上下文信息。最后在5个英文数据集和4个中文评论数据集上的实验结果表明,该文所提模型在准确率和F1上均表现最优。

宽卷积局部特征扩展的Transformer网络故障诊断模型

视觉Transformer网络的高精度诊断性能依赖于充分的训练数据,利用卷积网络在提取局部特征上的优势,构造能同时描述故障局部和全局特征的提取层,提高诊断模型的抗噪声干扰能力。首先,引入卷积网络模块将原始振动信号转换为Transformer网络可以直接接收的特征向量,提取故障局部特征,并通过增加卷积网络的感受野。然后,结合Transformer网络多头自注意力机制生成的全局信息,构建能同时描述故障局部和全局特征的特征向量。最后,在Transformer网络的预测层,利用高效通道注意力机制对特征向量的贡献度进行自动筛选。在西储大学(CWRU)轴承数据集上的故障诊断结果表明,在信噪比-4 dB的噪声干扰下,改进后的Transformer网络轴承故障诊断模型的准确率达90.21%,与原始Transformer模型相比,准确率提高了13.2%,在噪声环境下表现出优异的诊断性能。

Point Cloud Classification Using Content-Based Transformer via Clustering in Feature Space

Recently, there have been some attempts of Transformer in 3D point cloud classification. In order to reduce computations, most existing methods focus on local spatial attention,but ignore their content and fail to establish relationships between distant but relevant points. To overcome the limitation of local spatial attention, we propose a point content-based Transformer architecture, called PointConT for short. It exploits the locality of points in the feature space(content-based), which clusters the sampled points with similar features into the same class and computes the self-attention within each class, thus enabling an effective trade-off between capturing long-range dependencies and computational complexity. We further introduce an inception feature aggregator for point cloud classification, which uses parallel structures to aggregate high-frequency and low-frequency information in each branch separately. Extensive experiments show that our PointConT model achieves a remarkable performance on point cloud shape classification. Especially, our method exhibits 90.3% Top-1 accuracy on the hardest setting of ScanObjectN N. Source code of this paper is available at https://github.com/yahuiliu99/PointC onT.

RECURSIVE FILTERING RADON-AMBIGUITY TRANSFORM ALGORITHM FOR DETECTING MULTI-LFM SIGNALS

In multi-LFM signal condition, Radon-Ambiguity Transform (RAT) of the strong LFM component has strong suppression effect on that of the weak LFM component. A method named as Recursive Filtering RAT (RFRAT) algorithm is proposed for solving this problem. By fully using of the Maximum Likelihood (ML) estimation value of the frequency modulation rate got by RAT, RFRAT can detect the noisy multi-LFM signals out step by step. The merit of this new method is validated by an illustrative example in low Signal-to-Noise-Ratio (SNR) condition.

改进Transformer的高光谱图像地物分类方法——以黄河三角洲为例

高光谱技术已成为沿海湿地监测的主要手段,但传统高光谱分类方法通常存在特征提取不充分、同物异谱和场景碎片化等问题。针对这些问题,该文将Transformer用于高光谱分类,提出一种新的分类方法。该方法基于视觉自注意力模型(Vision Transformer,ViT),利用Non-local技术学习全局空间特征,扩大感受野解决提取判别特征不足的问题;同时,通过自适应跨层残差连接加强层间信息交换,解决信息损失的问题。选取NC16和NC13黄河三角洲湿地数据集作为实验数据,并将提出的方法与支持向量机(support vector machine,SVM)、一维卷积神经网络(one dimensional convolution neural network,1DCNN)、上下文深度卷积神经网络(contextual deep convolution neural network,CDCNN)、光谱空间残差网络(spectral-spatial residual network,SSRN)、混合光谱网络(hybrid spectral network,HybridSN)和ViT进行比较分析。结果表明,所提方法的总体精度(overall accuracy,OA)、平均精度(average accuracy,AA)和Kappa系数均有显著提高,OA分别达到96.24%和73.84%,AA分别达到83.42%和74.87%,Kappa分别达到94.80%和68.94%。

基于Local-Global-VIT细粒度分类算法的蝴蝶识别

【目的】准确鉴别蝴蝶种类,动态观测蝴蝶群落多样性变化对生境质量评估、生态环境恢复等方面具有重要意义。针对现有蝴蝶识别方法仅依靠整体特征,忽略了局部特征导致识别生态图像能力不足的问题,本研究旨在开发一种Local-Global-VIT细粒度分类算法的蝴蝶识别方法。【方法】本研究以5科200种共计25 279张蝴蝶图像为识别对象,采用多种数据增强方法扩充图像数据;通过视觉Transformer(vision transformer, VIT)层级结构及自注意力机制逐层选择局部令牌并保留至最后一层学习蝴蝶局部判别部位信息;聚合高层全局令牌消除复杂背景干扰;通过对比损失拉大类间距提高区分度。除此之外,使用合理的学习率调整策略和迁移学习方法,优化了模型收敛过程,在不增加参数量的情况下提高了性能。【结果】Local-Global-VIT算法在大规模细粒度公开数据集Butterfly-200上识别准确率达91.20%,较改进前提升了1.15%,比最优的一般害虫识别算法EfficientNet_b0和细粒度分类算法TransFG准确率分别高了1.83%和0.64%,F1分值分别提高了1.89%和0.88%。【结论】Local-Global-VIT算法以细粒度识别方式有效解决了蝴蝶类内差异大、类间差异小的分类难题,能准确地识别蝴蝶种类,有助于高效评估生境质量。

Localization method of subsynchronous oscillation source based on high-resolution time-frequency distribution image and CNN

The penetration of new energy sources such as wind power is increasing,which consequently increases the occurrence rate of subsynchronous oscillation events.However,existing subsynchronous oscillation source-identification methods primarily analyze fixed-mode oscillations and rarely consider time-varying features,such as frequency drift,caused by the random volatility of wind farms when oscillations occur.This paper proposes a subsynchronous oscillation sourcelocalization method that involves an enhanced short-time Fourier transform and a convolutional neural network(CNN).First,an enhanced STFT is performed to secure high-resolution time-frequency distribution(TFD)images from the measured data of the generation unit ports.Next,these TFD images are amalgamated to form a subsynchronous oscillation feature map that serves as input to the CNN to train the localization model.Ultimately,the trained CNN model realizes the online localization of subsynchronous oscillation sources.The effectiveness and accuracy of the proposed method are validated via multimachine system models simulating forced and natural oscillation events using the Power Systems Computer Aided Design platform.Test results show that the proposed method can localize subsynchronous oscillation sources online while considering unpredictable fluctuations in wind farms,thus providing a foundation for oscillation suppression in practical engineering scenarios.

基于局部Transformer的泰语分词和词性标注联合模型

泰语分词和词性标注任务二者之间存在高关联性,已有研究表明将分词和词性标注任务进行联合学习可以有效提升模型性能,为此,提出了一种针对泰语拼写和构词特点的分词和词性标注联合模型。针对泰语中字符构成音节,音节组成词语的特点,采用局部Transformer网络从音节序列中学习分词特征;考虑到词根和词缀等音节与词性的关联,将用于分词的音节特征融入词语序列特征,缓解未知词的词性标注特征缺失问题。在此基础上,模型采用线性分类层预测分词标签,采用线性条件随机场建模词性序列的依赖关系。在泰语数据集LST20上的试验结果表明,模型分词F1、词性标注微平均F1和宏平均F1分别达到96.33%、97.06%和85.98%,相较基线模型分别提升了0.33%、0.44%和0.12%。

基于改进Vision Transformer网络的农作物病害识别方法

基于DCNN模型的农作物病害识别方法在实验室环境下识别准确率高,但面对噪声时缺少鲁棒性.为了兼顾农作物病害识别的精度和鲁棒性,本文在标准ViT模型基础上加入增强分块序列化和掩码多头注意力,解决标准ViT模型缺乏局部归纳偏置和视觉特征序列的自注意力过于关注自身的问题.实验结果表明,本文的EPEMMSA-ViT模型对比标准ViT模型可以更高效的从零学习;当添加预训练权重训练网络时,EPEMMSA-ViT模型在数据增强的PlantVillage番茄子集上能够得到99.63%的分类准确率;在添加椒盐噪声的测试数据集上,对比ResNet50、DenseNet121、MobileNet和ConvNeXt的分类准确率分别提升了6.08%、9.78%、29.78%和12.41%;在添加均值模糊的测试数据集上,对比ResNet50、DenseNet121、MobileNet和ConvNeXt的分类准确率分别提升了18.92%、31.11%、20.37%和19.58%.

Swin-Transformer故障信息挖掘的海底观测网故障定位方法

海底观测网长期受海洋环境与人为因素影响,易使光电复合缆绝缘破损与海水接触形成电学故障点。如何准确地定位电学故障点,对提高海底观测网输电与信息传输的可靠性至关重要。首先根据海底观测网输电结构建立海底观测网输电模型,推导与模拟电学故障点传播至观测点的暂态电流,然后由连续小波变换提取暂态电流与故障点对应的内在关联特征量,最后通过Swin-Transformer神经网络挖掘内在关联特征量与故障距离的匹配关系来定位电学故障点。研究结果表明,在内在关联特征量样本测试集条件下,光电复合缆≤160 km的电学故障点定位误差小于400 m,可为长距离光电复合缆的海底观测网电学故障点定位提供参考。

基于SF-Transformer的智能教育平台短期电力负荷预测研究

建设智能教育平台是推动教育智能化的一个重要过程,但智能教育平台依赖的人工智能模型在训练过程中会消耗大量电力,因此,开展短期电力负荷预测对建设智能教育平台具有重要意义.针对在考虑多个属性开展短期电力负荷预测时,由于部分属性与电力负荷数据的相关性不强并且Transformer无法捕捉电力负荷数据的时间相关性,而导致电力负荷预测不够准确的问题,基于SR(Székely and Rizzo)距离相关系数、融合时间定位编码和Transformer,提出了一种短期电力负荷预测模型SF-Transformer.SF-Transformer通过SR距离相关系数对影响电力负荷数据的属性进行筛选,选择与电力负荷数据之间SR距离相关系数较大的属性.SF-Transformer采用一种全局时间编码与局部位置编码相结合的融合时间定位编码,有助于模型全面获取电力负荷数据的时间定位信息.在数据集上开展了实验,实验结果表明SF-Transformer与其他模型相比,在两种时长上进行电力负荷预测具有更低的均方根误差和平均绝对误差.

Online Capacitor Voltage Transformer Measurement Error State Evaluation Method Based on In-Phase Relationship and Abnormal Point Detection

The assessment of the measurement error status of online Capacitor Voltage Transformers (CVT) within the power grid is of profound significance to the equitable trade of electric energy and the secure operation of the power grid. This paper advances an online CVT error state evaluation method, anchored in the in-phase relationship and outlier detection. Initially, this method leverages the in-phase relationship to obviate the influence of primary side fluctuations in the grid on assessment accuracy. Subsequently, Principal Component Analysis (PCA) is employed to meticulously disentangle the error change information inherent in the CVT from the measured values and to compute statistics that delineate the error state. Finally, the Local Outlier Factor (LOF) is deployed to discern outliers in the statistics, with thresholds serving to appraise the CVT error state. Experimental results incontrovertibly demonstrate the efficacy of this method, showcasing its prowess in effecting online tracking of CVT error changes and conducting error state assessments. The discernible enhancements in reliability, accuracy, and sensitivity are manifest, with the assessment accuracy reaching an exemplary 0.01%.

Exploration and Practice of the Digital Transformation and Upgrading Paths for Civil Engineering Majors

This article examines the pathway to digital transformation and upgrading in undergraduate institutions,using the civil engineering program at Chongqing Three Gorges University as a case study,focusing on six key areas:developing a high-quality digital talent training program for civil engineering;assembling diverse resources to create a digital,multi-scenario open learning environment that encompasses teaching,research,and practical training for civil engineering undergraduates;piloting innovative digital teaching models for civil engineering undergraduates;crafting a new model for digital resource provision,utilizing self-developed and specialized resources;devising assessment methods and ongoing improvement strategies based on the achievement of students’digital competencies;and devising a new,three-dimensional,multi-modal teaching evaluation system through intelligent data capture and analysis.

Optimization Analysis of Human Resource Management Strategies of Nantong University,a Local Comprehensive University in China

This study analyzes the current situation and challenges of Nantong University amid the rapid development of higher education,addressing aspects such as faculty structure,administrative management efficiency,and promotion mechanisms.Through questionnaire surveys and field research,coupled with SWOT analysis and various theoretical methods,a comprehensive analysis was conducted on Nantong University’s internal and external environment.The study proposes an improved comprehensive development strategy and implementation measures,which include updating management concepts,strengthening professional training,optimizing employment conditions,and improving salary systems.Ultimately,this study provides practical strategies for human resource management and the transformational development of Nantong University and other local comprehensive universities.

基于Transformer与局部特征融合的轨道紧固件缺陷检测方法

为解决传统人工巡检轨道交通线路存在的效率低和有安全隐患等问题,提出一种基于Transformer与局部特征融合的轨道紧固件缺陷检测方法。构建轨道紧固件缺陷检测模型,将Transformer与局部特征模块融合,整合局部信息,进而提取轨道紧固件缺陷特征;同时,采用数据增强的方法对轨道紧固件缺陷样本进行数据扩增,扩充数据集,验证所建模型的检测效果。实验结果表明,相较于传统方法,文章提出的方法在识别轨道紧固件缺失和损坏两类缺陷方面的精度和平均准确率均有所提升,在不同的轨道线路实验环境下也表现出良好的检测效果。

一种基于Transformer模型的特征增强算法及其应用研究

Transformer模型在自动语音识别(ASR)任务中展现出优秀的性能,但在特征提取方面存在两个问题:一是模型集中于全局特征交互信息提取,忽略了其他有用的特征信息,如局部特征交互信息;二是模型对低层特征交互信息的利用不够充分。为了解决这两个问题,提出了卷积线性映射(CMLP)模块以强化局部特征交互,并设计低层特征融合(LF)模块来融合高低层特征。通过整合这些模块,构建了CLformer模型。在两个中文普通话数据集(Aishell-1和HKUST)上进行实验,结果表明,CLformer显著提升了模型性能,在Aishell-1上较基线提高0.3%,在HKUST上提高0.5%。

基于Swin Transformer和图形推理的结直肠息肉分割方法

针对结直肠息肉图像分割中病灶区域尺度变化大、边缘模糊以及息肉与正常组织对比度低等问题,导致病变区域分割精度低和分割边界存在伪影,提出一种基于Swin Transformer和图形推理的自适应网络.该网络一是利用Swin Transformer编码器逐层提取输入图像的全局上下文信息,弱化背景信息干扰,多尺度分析病变区域的显著性特点.二是提出全局与局部特征交互模块增强网络对复杂病灶的空间感知能力,突出待分割目标的关键位置信息,精准定位目标.三是通过区域引导图推理模块以图循环递推的方式挖掘先验信息之间的高阶显性关系,促进图间信息传递.四是设计面向边缘细节的边缘约束图推理模块,整合边缘细节,改善分割效果,提高分割精度.在CVC-ClinicDB、Kvasir、CVC-ColonDB和ETIS数据集上进行实验,其Dice系数分别为0.939,0.926,0.810和0.788,平均交并比分别为0.889,0.879,0.731和0.710,分割性能优于现有方法.仿真实验结果表明,对于形态结构复杂、对比度低和边缘模糊的结直肠息肉图像均有较高的分割精度.

基于局部Radon-Ambiguity变换的ARM检测技术

针对反辐射导弹(anti-radiation missile,ARM)雷达回波信号的调频率特性,提出了基于局部Radon-Ambiguity变换(RAT)的ARM检测方法。该方法对导弹的初速度不敏感,也不需要对导弹加速度进行多通道相位补偿预处理。通过将RAT局部化,进一步缩小了检测区域和运算量,提高了检测效率。检测性能分析和仿真实验表明,该方法能够在较低信噪比环境下有效地检测ARM目标信号。