期刊文献+
任意字段
题名或关键词
题名
关键词
摘要
作者
第一作者
机构
刊名
分类号
参考文献
作者简介
基金资助
栏目信息
共找到5,394篇文章
< 1 2 … 250 >
每页显示 20 50 100
已选0条
引用分析
参考文献
引证文献
 统计分析
检索结果
已选文献
显示方式：
Practical implementation of Hilbert-Huang Transform algorithm 认领 引用 被引量：53
1
作者 黄大吉 赵进平 苏纪兰 《Acta Oceanologica Sinica》 SCIE CAS 2003年第1期1-14,共14页
Hilbert-Huang Transform (HHT) is a newly developed powerful method for nonlinear and non-stationary time series analysis. The empirical mode decomposition is the key part of HHT, while its algorithm was protected by N... Hilbert-Huang Transform (HHT) is a newly developed powerful method for nonlinear and non-stationary time series analysis. The empirical mode decomposition is the key part of HHT, while its algorithm was protected by NASA as a US patent, which limits the wide application among the scientific community. Two approaches, mirror periodic and extrema extending methods, have been developed for handling the end effects of empirical mode decomposition. The implementation of the HHT is realized in detail to widen the application. The detailed comparison of the results from two methods with that from Huang et al. (1998, 1999), and the comparison between two methods are presented. Generally, both methods reproduce faithful results as those of Huang et al. For mirror periodic method (MPM), the data are extended once forever. Ideally, it is a way for handling the end effects of the HHT, especially for the signal that has symmetric waveform. The extrema extending method (EEM) behaves as good as MPM, and it is better than MPM for the signal that has strong asymmetric waveform. However, it has to perform extrema envelope extending in every shifting process. 展开更多
关键词 Hilbert-Huang Transform (HHT) signal processing Empirical Mode Decomposition (EMD)
暂未订购 下载PDF
Shaking table test on the seismic failure mechanism of a deposit slope with a weak interlayer:Insights from Hilbert-Huang Transform analysis 认领 引用
2
作者 WEN Hao HE Yifan +3 位作者 XU Yinchong DONG Xiaoqiang DING Xuanming LI Xunfeng 《Journal of Mountain Science》 SCIE CSCD 2026年第3期1221-1233,共13页
Earthquakes are critical triggers for slope instability.While extensive research has been conducted on slope failure modes under seismic loading,the identification of sliding surface propagation and coalescence remain... Earthquakes are critical triggers for slope instability.While extensive research has been conducted on slope failure modes under seismic loading,the identification of sliding surface propagation and coalescence remains insufficiently explored.This study investigates the dynamic response of a deposit slope containing a weak interlayer through large-scale shaking table tests.The propagation process of the sliding surface was identified using the Hilbert-Huang transform and marginal spectrum analysis.Under seismic excitation,sliding occurs along the interface between the overburden and the weak interlayer,leading to sudden landslide events.Differential vibrations at the overburden-weak interlayer-bedrock interfaces are identified as a primary mechanism driving landslide initiation.As input acceleration increases,these interfacial vibration contrasts intensify,and the acceleration amplification effect within the overburden becomes markedly pronounced.Following landslide occurrence,the vibration differences across interfaces decrease sharply.In the time-frequency domain,seismic waves transmitted through the weak interlayer exhibit amplified low-frequency components.Marginal spectrum analysis of seismic energy evolution within the slope reveals that energy attenuation in the 19-22 Hz frequency band correlates with landslide occurrence,while attenuation in the 9-11 Hz band serves as an indicator for sliding surface propagation and coalescence.For seismic design of deposit slopes with weak interlayers,particular attention should be given to the increased seismic inertial forces in the overburden layer and the detrimental effects of low-frequency wave components on sliding surface development. 展开更多
关键词 Earthquake Shaking table test Dynamic characteristic Hilbert-Huang Transform Marginal spectra Zheduoshan Tunnel
暂未订购 下载PDF
基于Transformer的DETR目标检测算法综述 认领 引用 被引量：7
3
作者 李沂杨 陆声链 +1 位作者 王继杰 陈明 《计算机工程》 CAS CSCD 北大核心 2026年第4期62-81,共20页
在目标检测领域,卷积神经网络(CNN)凭借其优异的准确性和可扩展性,长期主导着相关研究,并获得了学术界的广泛认可。在此框架下,先后涌现出基于区域的卷积神经网络(R-CNN)系列(如Fast R-CNN、Faster R-CNN)与YOLO(You Only Look Once)系... 在目标检测领域,卷积神经网络(CNN)凭借其优异的准确性和可扩展性,长期主导着相关研究,并获得了学术界的广泛认可。在此框架下,先后涌现出基于区域的卷积神经网络(R-CNN)系列(如Fast R-CNN、Faster R-CNN)与YOLO(You Only Look Once)系列等多个代表性模型。随着Transformer在自然语言处理领域的成功,研究者开始探索将其用于计算机视觉领域,由此产生了视觉Transformer(ViT)和Swin Transformer等视觉骨干网络。Facebook团队为减少目标检测任务中的先验知识和后处理,在2020年推出了一种端到端目标检测算法——基于Transformer的DETR(DEtection TRansformer)。尽管DETR在目标检测领域展现出潜力,但也存在收敛速度慢、准确性较差、目标查询的物理意义不明确等缺点。这促使研究者对该算法开展了进一步的研究和改进。本研究旨在归纳总结针对DETR的改进探索,并分析它们的优势与不足,同时对利用DETR开展的前沿研究和细分应用领域进行概括,最后给出DETR在计算机视觉领域的未来展望。 展开更多
关键词 计算机视觉 目标检测 DETR算法 视觉Transformer 图像分割
暂未订购 下载PDF
Improved Cuckoo Search Algorithm for Engineering Optimization Problems 认领 引用
4
作者 Shao-Qiang Ye Azlan Mohd Zain Yusliza Yusoff 《Computers, Materials & Continua》 SCIE EI 2026年第4期1607-1631,共25页
Engineering optimization problems are often characterized by high dimensionality,constraints,and complex,multimodal landscapes.Traditional deterministic methods frequently struggle under such conditions,prompting incr... Engineering optimization problems are often characterized by high dimensionality,constraints,and complex,multimodal landscapes.Traditional deterministic methods frequently struggle under such conditions,prompting increased interest in swarm intelligence algorithms.Among these,the Cuckoo Search(CS)algorithm stands out for its promising global search capabilities.However,it often suffers from premature convergence when tackling complex problems.To address this limitation,this paper proposes a Grouped Dynamic Adaptive CS(GDACS)algorithm.Theenhancements incorporated intoGDACS can be summarized into two key aspects.Firstly,a chaotic map is employed to generate initial solutions,leveraging the inherent randomness of chaotic sequences to ensure a more uniform distribution across the search space and enhance population diversity from the outset.Secondly,Cauchy and Levy strategies replace the standard CS population update.This strategy involves evaluating the fitness of candidate solutions to dynamically group the population based on performance.Different step-size adaptation strategies are then applied to distinct groups,enabling an adaptive search mechanism that balances exploration and exploitation.Experiments were conducted on six benchmark functions and four constrained engineering design problems,and the results indicate that the proposed GDACS achieves good search efficiency and produces more accurate optimization results compared with other state-of-the-art algorithms. 展开更多
关键词 Cuckoo search algorithm chaotic transformation population division adaptive update strategy Cauchy distribution
暂未订购 下载PDF
基于Transformer的地下煤炭气化产物的预测模型 认领 引用 被引量：1
5
作者 袁淑霞 李铮 +2 位作者 王创业 吴松 张佳锋 《煤炭技术》 CAS 2026年第5期31-36,共6页
为推动煤炭能源绿色低碳转型,解决煤炭地下气化(UCG)产物预测中多输入多输出、注入产出时间滞后、气化反应与组分约束等难点,建立了基于Transformer算法的煤炭地下气化产物预测模型。结合UCG三区反应机理与Transformer算法的架构特征,... 为推动煤炭能源绿色低碳转型,解决煤炭地下气化(UCG)产物预测中多输入多输出、注入产出时间滞后、气化反应与组分约束等难点,建立了基于Transformer算法的煤炭地下气化产物预测模型。结合UCG三区反应机理与Transformer算法的架构特征,构建时间滞后特征、梯度特征、对数变换特征及经验模态分解(EMD)的算法模型。模型整合了物理约束、多任务学习及贝叶斯优化机制。模型预测结果显示:主要气体和温度的决定系数均超0.96,整体均方根误差与平均绝对误差较小,无显著系统误差,可有效捕捉UCG产物变化规律。 展开更多
关键词 煤炭地下气化 Transformer算法 多任务学习 贝叶斯优化 产物预测
暂未订购 下载PDF
基于ALA优化VMD和Transformer-BiGRU的锂离子电池RUL预测 认领 引用 被引量：1
6
作者 闫来清 梁朝洋 +4 位作者 韩磊 李沁春 蔡钰濠 魏嘉轩 许志超 《电源技术》 CAS 北大核心 2026年第2期277-289,共13页
准确预测电池剩余使用寿命(RUL)对保障电池管理系统的安全运行至关重要。然而,电池容量再生效应和测量误差会影响RUL预测精度,传统经验模态分解(EMD)虽能分解信号,却存在模态混叠问题;此外,现有方法中变分模态分解、预测模型核心参数大... 准确预测电池剩余使用寿命(RUL)对保障电池管理系统的安全运行至关重要。然而,电池容量再生效应和测量误差会影响RUL预测精度,传统经验模态分解(EMD)虽能分解信号,却存在模态混叠问题;此外,现有方法中变分模态分解、预测模型核心参数大多依赖手动设置,不仅耗时费力且预测效果受主观因素影响较大。针对上述难题,提出了一种基于ALA优化VMD和Transformer-BiGRU的预测方法。利用人工旅鼠算法(ALA)对变分模态分解(VMD)的模态分解层数与惩罚因子进行优化,通过VMD将电池容量数据分解为不同频率的模态分量,减小容量再生效应和测量误差的影响。构建了结合注意力机制(Transformer)和双向门控循环单元(BiGRU)的预测模型,并通过海星优化算法(SFOA)优化模型结构和关键参数,将各分量的预测结果加和得到最终预测值。基于MATLAB平台在两组公开数据集上验证了所提模型的有效性,并与其他预测方法进行对比。结果表明,所提方法的均方根误差(RMSE)均低于0.5%,预测精度较高。 展开更多
关键词 锂离子电池 人工旅鼠算法 海星优化算法 变分模态分解 双向门控循环单元 Transformer模型
暂未订购 下载PDF
基于Transformer网络的线性密码攻击方法研究 认领 引用
7
作者 张磊 李子凡 +2 位作者 李国元 肖超恩 王建新 《计算机工程与科学》 CSCD 北大核心 2026年第7期1204-1210,共7页
针对传统的分组密码分析方法需要复杂的理论推导和大量数据统计,并且数据复杂度和时间复杂度相对较高等问题,提出了一种基于Transformer网络的线性区分器模型,并设计了一种线性密码攻击方法,该方法以DES密码算法为例进行了模型参数分析... 针对传统的分组密码分析方法需要复杂的理论推导和大量数据统计,并且数据复杂度和时间复杂度相对较高等问题,提出了一种基于Transformer网络的线性区分器模型,并设计了一种线性密码攻击方法,该方法以DES密码算法为例进行了模型参数分析和3轮单密钥位的恢复攻击实验。攻击实验结果表明,基于Transformer的线性区分器模型准确率达到了69.75%,比ResNet模型提高了2.52个百分点。 展开更多
关键词 Transformer网络 线性区分器 DES密码算法
暂未订购 下载PDF
基于Transformer的单目标跟踪研究进展 认领 引用
8
作者 张大伟 徐东升 +3 位作者 余哲晨 蒋凯薇 田卫刚 郑忠龙 《模式识别与人工智能》 EI CSCD 北大核心 2026年第4期348-378,共31页
单目标跟踪(Single Object Tracking,SOT)是计算机视觉领域的核心任务之一.然而,在复杂场景中,传统的相关滤波和孪生网络架构已难以满足更高的精度与鲁棒性要求.近年来,Transformer架构凭借其卓越的全局建模能力,在SOT中表现出显著优势... 单目标跟踪(Single Object Tracking,SOT)是计算机视觉领域的核心任务之一.然而,在复杂场景中,传统的相关滤波和孪生网络架构已难以满足更高的精度与鲁棒性要求.近年来,Transformer架构凭借其卓越的全局建模能力,在SOT中表现出显著优势.因此,文中系统梳理基于Transformer的SOT研究进展.依据整体结构的设计,将现有算法分为双流双阶段跟踪算法与单流单阶段跟踪算法,并深入归纳分析其中具有代表性的跟踪算法,同时总结轻量化Transformer跟踪方法的研究现状.此外,进一步探讨基于Mamba的跟踪方法和通用模型等前沿热点,它们在模型效率与通用性方面展现出良好潜力.在多个主流数据集上对不同Transformer跟踪算法的性能进行全面评估和对比分析.最后,展望SOT在模型轻量化、多模态融合、长时跟踪、大模型驱动跟踪等方面的研究趋势,期望为单目标跟踪的研究与发展提供参考. 展开更多
关键词 计算机视觉 Transformer 双流双阶段跟踪算法 单流单阶段跟踪算法
暂未订购 下载PDF
基于轻量化Swin Transformer的资源自适应语义压缩策略 认领 引用
9
作者 郑飞 吴东颖 +5 位作者 李世超 仇洪冰 邵苏杰 喻鹏 丰雷 赵继龙 《通信学报》 EI CSCD 北大核心 2026年第6期146-159,共14页
在面向多终端的下行语义通信场景中,由于终端算力的差异性,采用统一的语义压缩比提取语义信息,会造成低算力终端难以完成语义解码、高算力终端的算力资源未充分利用和语义数据不够精练的问题。为解决上述问题,针对有限的终端算力资源和... 在面向多终端的下行语义通信场景中,由于终端算力的差异性,采用统一的语义压缩比提取语义信息,会造成低算力终端难以完成语义解码、高算力终端的算力资源未充分利用和语义数据不够精练的问题。为解决上述问题,针对有限的终端算力资源和链路带宽,提出了一种基于轻量化Swin Transformer的资源自适应语义压缩策略,采用门控网络和稀疏注意力机制改进了语义编码器结构,为终端定制差异化语义压缩比,以最小化系统总能耗为优化目标,构建联合算力和带宽分配模型,并采用近端策略优化(PPO)算法求解优化问题。仿真结果表明,与固定语义压缩比方案相比,所提策略的系统总能耗降低了39%。 展开更多
关键词 语义通信 Swin Transformer 稀疏注意力机制 近端策略优化算法
暂未订购 下载PDF
考虑INGO优化算法与VMD分解技术的TCN-Transformer的光伏出力预测方法研究 认领 引用
10
作者 郑非凡 许野 +2 位作者 王旭 孟亦康 李忠艳 《太阳能学报》 EI CAS CSCD 北大核心 2026年第5期741-754,共14页
提出基于改进北方苍鹰优化算法(INGO)和变分模态分解技术(VMD)的TCN-Transformer光伏出力组合预测模型。首先,应用动态时间规整算法(DTW)计算待预测日与历史日的气象序列之间的累积距离,生成与待预测日天气特征相似的历史日集合;其次,... 提出基于改进北方苍鹰优化算法(INGO)和变分模态分解技术(VMD)的TCN-Transformer光伏出力组合预测模型。首先,应用动态时间规整算法(DTW)计算待预测日与历史日的气象序列之间的累积距离,生成与待预测日天气特征相似的历史日集合;其次,联合使用Logistic混沌映射、等级制度、Levy方法和非线性因子对NGO进行改进,妥善解决传统的NGO易陷入局部最优的缺陷;然后,使用INGO和VMD的全新组合将待预测日的原始发电量序列分解为不同的本征模态函数(IMF);最后,将气象数据和IMF分量分别作为输入和输出变量,训练生成基于INGO的TCN-Transformer组合预测模型。依托云南某光伏电站实测应用数据,验证所提方法的优越性,在晴天和阴雨天气条件下均具有较高的预测精度。 展开更多
关键词 光伏发电 预测模型 优化算法 变分模态分解 时间卷积网络 Transformer模型
暂未订购 下载PDF
混合注意力嵌入Transformer的岩心CT图像超分辨率重建 认领 引用
11
作者 李艳辉 杨延坤 《中国石油大学学报（自然科学版）》 EI CAS CSCD 北大核心 2026年第2期195-201,共7页
利用CT(computed tomography)技术对岩石进行成像时,由于成像视野与成像分辨率成反比,为获得更多的岩石信息需要更大的成像视野,得到的岩石图像分辨率较低。针对此问题提出一种基于混合注意力机制与Transformer的岩心CT图像超分辨率重... 利用CT(computed tomography)技术对岩石进行成像时,由于成像视野与成像分辨率成反比,为获得更多的岩石信息需要更大的成像视野,得到的岩石图像分辨率较低。针对此问题提出一种基于混合注意力机制与Transformer的岩心CT图像超分辨率重建网络,该模型特征提取部分结合十字交叉注意力机制与窗口间的重叠交叉注意力机制,提升网络对岩心CT图像中特征信息的提取和像素点的利用能力,恢复图像清晰细节。结果表明,在尺度因子为4和8倍岩心CT图像重建下,与其他先进模型相比,本文中方法在岩心CT图像超分辨率重建任务中模型的性能指标得到了提升,峰值信噪比分别提高0.15和0.21 dB。 展开更多
关键词 超分辨率 Transformer算法 注意力机制 岩心CT图像 卷积神经网络
暂未订购 下载PDF
自适应Transformer-LSTM的滚动轴承故障预测方法研究 认领 引用
12
作者 董辛旻 职帅轩 崔富源 《重庆理工大学学报（自然科学）》 CAS 北大核心 2026年第2期96-104,共9页
针对滚动轴承故障预测中,模型预测精度低、参数难以确定等问题,提出一种自适应Transformer编码器与长短期记忆神经网络(long short-term memory, LSTM)的滚动轴承故障预测方法。提取轴承全周期振动信号24维时频特征,并基于信息熵、鲁棒... 针对滚动轴承故障预测中,模型预测精度低、参数难以确定等问题,提出一种自适应Transformer编码器与长短期记忆神经网络(long short-term memory, LSTM)的滚动轴承故障预测方法。提取轴承全周期振动信号24维时频特征,并基于信息熵、鲁棒性和相关性构建综合指标对特征进行筛选。采用核主元分析将优选特征进行融合,构建能够反应轴承退化状态的退化指标。综合时间序列全局特性与长短期特性,构建Transformer编码器与LSTM结合的故障预测模型,并引入极光优化算法(polar lights optimization, PLO)进行参数自适应寻优,应用XJTU-SY与IMS滚动轴承数据集进行方法分析验证。结果表明:所提模型在轴承故障预测中表现出良好效果,相较于经典的时域卷积网络(temporal convolutional network, TCN)、双向门控循环神经网络(gate recurrent unit, GRU)等,预测精度有较大提升。 展开更多
关键词 滚动轴承 故障预测 数据驱动 Transformer 优化算法
暂未订购 下载PDF
基于MSWOA-Transformer-LSTM混合神经网络的沿海高桩码头结构健康监测数据预测方法研究 认领 引用
13
作者 刘红彪 王之祥 +4 位作者 刘慧芳 罗德通 齐方利 张路刚 薛德清 《振动与冲击》 EI CSCD 北大核心 2026年第11期144-154,共11页
结构健康监测(structural health monitoring,SHM)系统在长期运行中会产生海量监测数据,为了实现码头结构安全状态的快速预测,需建立高效的数据预测模型。针对现有的结构健康监测数据预测模型训练效率低、模型预测精度不足、泛化能力弱... 结构健康监测(structural health monitoring,SHM)系统在长期运行中会产生海量监测数据,为了实现码头结构安全状态的快速预测,需建立高效的数据预测模型。针对现有的结构健康监测数据预测模型训练效率低、模型预测精度不足、泛化能力弱及易陷入局部最优等问题,提出了一种基于Transformer网络、长短期记忆网络(long short-term memory,LSTM)和基于混合策略改进的鲸鱼优化算法(multi-strategy improved whale optimization algorithm,MSWOA)相结合的监测数据预测分析模型,即MSWOA-Transformer-LSTM(MTL)模型,并针对天津港某高桩码头健康监测系统实测的结构应变、保护电位和温度等多源数据进行了分析和预测。结果表明,因MSWOA方法的引入,MTL模型的寻优精度和收敛速度明显提升。同时,通过与Transformer-GRU(gated recurrent unit)、Transformer-LSTM和WOA-Transformer-LSTM(WTL)等神经网络模型的预测结果对比可知,MTL模型在预测任务中均表现出更高的精度和更好的稳定性。 展开更多
关键词 结构健康监测(SHM) 混合策略改进的鲸鱼优化算法(MSWOA) 长短期记忆网络(LSTM) Transformer 高桩码头 数据预测
暂未订购 下载PDF
A Full-Newton Step Feasible Interior-Point Algorithm for the Special Weighted Linear Complementarity Problems Based on Algebraic Equivalent Transformation 认领 引用
14
作者 Jing GE Mingwang ZHANG Panjie TIAN 《Journal of Mathematical Research with Applications》 CSCD 2025年第4期555-568,共14页
In this paper,we propose a new full-Newton step feasible interior-point algorithm for the special weighted linear complementarity problems.The proposed algorithm employs the technique of algebraic equivalent transform... In this paper,we propose a new full-Newton step feasible interior-point algorithm for the special weighted linear complementarity problems.The proposed algorithm employs the technique of algebraic equivalent transformation to derive the search direction.It is shown that the proximity measure reduces quadratically at each iteration.Moreover,the iteration bound of the algorithm is as good as the best-known polynomial complexity for these types of problems.Furthermore,numerical results are presented to show the efficiency of the proposed algorithm. 展开更多
关键词 interior-point algorithm weighted linear complementarity problem algebraic equivalent transformation search direction iteration complexity
暂未订购 下载PDF
RetinexWT: Retinex-Based Low-Light Enhancement Method Combining Wavelet Transform 认领 引用
15
作者 Hongji Chen Jianxun Zhang +2 位作者 Tianze Yu Yingzhu Zeng Huan Zeng 《Computers, Materials & Continua》 SCIE EI 2026年第2期2113-2132,共20页
Low-light image enhancement aims to improve the visibility of severely degraded images captured under insufficient illumination,alleviating the adverse effects of illumination degradation on image quality.Traditional ... Low-light image enhancement aims to improve the visibility of severely degraded images captured under insufficient illumination,alleviating the adverse effects of illumination degradation on image quality.Traditional Retinex-based approaches,inspired by human visual perception of brightness and color,decompose an image into illumination and reflectance components to restore fine details.However,their limited capacity for handling noise and complex lighting conditions often leads to distortions and artifacts in the enhanced results,particularly under extreme low-light scenarios.Although deep learning methods built upon Retinex theory have recently advanced the field,most still suffer frominsufficient interpretability and sub-optimal enhancement performance.This paper presents RetinexWT,a novel framework that tightly integrates classical Retinex theory with modern deep learning.Following Retinex principles,RetinexWT employs wavelet transforms to estimate illumination maps for brightness adjustment.A detail-recovery module that synergistically combines Vision Transformer(ViT)and wavelet transforms is then introduced to guide the restoration of lost details,thereby improving overall image quality.Within the framework,wavelet decomposition splits input features into high-frequency and low-frequency components,enabling scale-specific processing of global illumination/color cues and fine textures.Furthermore,a gating mechanism selectively fuses down-sampled and up-sampled features,while an attention-based fusion strategy enhances model interpretability.Extensive experiments on the LOL dataset demonstrate that RetinexWT surpasses existing Retinex-oriented deeplearning methods,achieving an average Peak Signal-to-Noise Ratio(PSNR)improvement of 0.22 dB over the current StateOfTheArt(SOTA),thereby confirming its superiority in low-light image enhancement.Code is available at http://gffzz188fe103f8f1460asxpq69kb5bcpn6bww.ffgz.tsg.suse.edu.cn/CHEN-hJ516/RetinexWT(accessed on 14 October 2025). 展开更多
关键词 Low-light image enhancement retinex algorithm wavelet transform vision transformer
暂未订购 下载PDF
基于改进Transformer的连铸机零段基座状态识别研究 认领 引用 被引量：2
16
作者 李欣 但斌斌 +3 位作者 刘洋 容芷君 都李平 肖林伟 《武汉科技大学学报》 CAS 北大核心 2026年第4期295-306,共12页
针对连铸机零段基座状态难以实时、准确识别的问题,提出一种基于改进鲸鱼优化算法优化的融合时序卷积Transformer的零段基座状态识别方法。首先,提取基座压力时序数据的时域和频域特征,并采用核主成分分析进行非线性降维;在此基础上,利... 针对连铸机零段基座状态难以实时、准确识别的问题,提出一种基于改进鲸鱼优化算法优化的融合时序卷积Transformer的零段基座状态识别方法。首先,提取基座压力时序数据的时域和频域特征,并采用核主成分分析进行非线性降维;在此基础上,利用熵权法对特征进行加权重构,构建加权多域特征集。其次,在Transformer模型中引入时序卷积层和最大池化层,以增强模型对局部时序特征的提取能力,并降低模型参数量降低过拟合风险。最后,通过引入非线性收敛因子、自适应权重和柯西扰动策略改进鲸鱼优化算法,并利用改进算法对CTransformer模型关键超参数进行寻优,建立零段基座状态识别模型。以某炼钢厂现场采集的基座压力信号进行实验验证,结果表明,所提方法在测试集上的平均识别准确率为98.93%,分别较SVM、CNN、GRU和Transformer模型提高19.32、16.38、13.31和8.26个百分点,可为连铸机零段基座状态在线监测提供参考。 展开更多
关键词 连铸机零段基座 状态识别 时序卷积 鲸鱼优化算法 Transformer
暂未订购 下载PDF
Seismic Data Reconstruction and Noise Suppression Based on an Improved Weighted Fast Projection onto Convex Sets Algorithm 认领 引用
17
作者 Cheng Gao-peng Zhang Hua +3 位作者 Li Hong-xing Song Yu Yue Ming Zhang Kai-dong 《Applied Geophysics》 SCIE CSCD 2026年第1期336-351,433,434,共16页
In eld seismic data acquisition,seismic traces are often aected by substantial data gaps and strong noise interference due to environmental and instrumental factors,thus degrading the resolution and signalto-noise rat... In eld seismic data acquisition,seismic traces are often aected by substantial data gaps and strong noise interference due to environmental and instrumental factors,thus degrading the resolution and signalto-noise ratio(SNR)of the seismic profiles.Effective seismic data reconstruction and noise suppression techniques are therefore essential to recover missing signals and improve data quality.In this study,a fast projection onto convex sets(FPOCS)algorithm is proposed by incorporating an inertial parameter that involves a linear combination of the two preceding iterations based on the traditional projection onto convex sets(POCS)algorithm.Then,a weighting factor is introduced to achieve simultaneous data reconstruction and noise suppression using the weighted fast projection onto convex sets(WFPOCS)algorithm.To further suppress residual random noise in the updated solution,an optimization strategy is adopted by swapping the order of the iterative hard thresholding operator and the projection operator.The nal algorithm,termed the improved weighted fast projection onto convex sets(IWFPOCS),achieves high-efciency reconstruction and effective noise suppression.Compared with WFPOCS,the proposed method maintains fast reconstruction speed while demonstrating superior denoising performance on irregularly missing and noisy datasets.Field data experiments conrm that the proposed method signicantly improves the SNR and resolution of seismic data,oering strong practical potential for subsequent processing and interpretation. 展开更多
关键词 Seismic data reconstruction Fast projection onto convex set algorithm Curvelet transform Weighted fast projection onto convex set algorithm Noise suppression
暂未订购 下载PDF
A Novel Hybrid Evolutionary Transformer-Long Short-Term Memory Model for Unified Anomaly Detection in IoT and Cyber-Physical Networks 认领 引用
18
作者 Pardis Sadatian Moghaddam Mahyar Mahmoudi +3 位作者 Nuria Serrano Francisco Hernando-Gallego Diego Martín José Vicente Álvarez-Bravo 《Computers, Materials & Continua》 SCIE EI 2026年第9期2068-2108,共41页
The rapid proliferation of the Internet of Things(IoT)and cyber-physical systems(CPS)within critical infrastructure sectors has significantly expanded the attack surface for advanced and stealthy cyber threats.Since t... The rapid proliferation of the Internet of Things(IoT)and cyber-physical systems(CPS)within critical infrastructure sectors has significantly expanded the attack surface for advanced and stealthy cyber threats.Since these systems increasingly rely on real-time data exchange and autonomous control,developing intelligent,scalable,and adaptive anomaly detection mechanisms has become a pressing requirement.This paper proposes a novel hybrid framework,evolutionary-transformer-long short-term memory(Evo-Transformer-LSTM),that integrates the temporal modeling capability of LSTM networks,the global attention mechanism of Transformer encoders,and the optimization power of the improved chimp optimization algorithm(IChOA)for hyper-parameter tuning.In the proposed architecture,the Transformer encoder extracts high-level contextual patterns from traffic sequences,while the LSTM component captures local temporal dependencies.The framework is rigorously evaluated on four benchmark datasets from the Canadian Institute for Cybersecurity(CIC):CIC-IDS-2017,CSE-CIC-IDS-2018,CIC-IoT-DIAD(2024),and CIC-IoV(2024).Comparative experiments are conducted against several state-of-the-art baselines,including transformer,LSTM,bidirectional encoder representations from transformers(BERT),deep reinforcement learning(DRL),convolutional neural network(CNN),k-nearest neighbors(KNN),and random forest(RF)classifiers.Results show that the proposed Evo-Transformer-LSTM achieves up to 98.25%accuracy,an F1-score of 97.91%,and an area under the curve(AUC)of 99.36%on CIC-IDS 2017,while maintaining above 96%accuracy and 98%AUC even on the more challenging CIC-IoV 2024 dataset,consistently surpassing all baseline models.In addition,statistical significance tests confirm the superiority of the proposed approach.In conclusion,Evo-Transformer-LSTM offers a unified,scalable,and robust solution for anomaly detection in modern IoT and CPS infrastructures,with potential for real-world deployment in security-sensitive domains. 展开更多
关键词 Internet of Things cyber-physical systems intrusion detection datasets transformer long short-term memory chimp optimization algorithm
暂未订购 下载PDF
Enhanced multiscale quantum approximate optimization algorithm in multibody combinatorial optimization problems 认领 引用
19
作者 Lei-Lei Chen Ping Zou Ya-Fei Yu 《Chinese Physics B》 SCIE EI CAS CSCD 2026年第7期424-434,共11页
At present,the quantum approximate optimization algorithm(QAOA)faces scalability challenges in high-dimensional combinatorial optimization problems due to exponentially growing computational costs and reachability def... At present,the quantum approximate optimization algorithm(QAOA)faces scalability challenges in high-dimensional combinatorial optimization problems due to exponentially growing computational costs and reachability deficits for noisy intermediate-scale quantum(NISQ)devices.This study focuses on the multiscale quantum approximate optimization algorithm(MQAOA),which integrates renormalization group(RG)transformations with QAOA to address these limitations.Based on the connections between the variables in the problem to be solved,the weighted maximal matching method is employed to generate a variable partitioning strategy guiding the RG transformation.This approach not only extends the applicability of MQAOA to satisfiability(SAT)problems—including those with three-body and higher-order interactions in the problem Hamiltonian—but also eliminates the algorithm's sensitivity to problem density.Validations conducted on quantum simulators show that,after running two-round MQAOA,its capability is enhanced to identify optimal solutions with approximately 97%success probability as defined by the ground-state overlap for Max-2-SAT problems(78%success probability for Max-3-SAT problems).The results confirm the feasibility of MQAOA and establish it as a resource-efficient framework for complex combinatorial optimization problems,providing a pathway for NISQ-era deployment. 展开更多
关键词 quantum algorithm parameterized quantum circuit renormalization group transformation combinatorial optimization Max-SAT problems
暂未订购 下载PDF
融合鲸鱼算法与Transformer的乌溪江支墩坝变形预测模型研究 认领 引用 被引量：3
20
作者 琚海荣 梁昱晨 +1 位作者 晁阳 齐慧君 《人民珠江》 2026年第4期43-51,共9页
针对老旧支墩坝变形机理复杂、传统预测模型精度不足的问题,提出了基于鲸鱼优化算法(Whale OptimizationAlgorithm,WOA)改进的Transformer变形预测模型。该方法利用WOA对Transformer模型的核心超参数进行全局寻优,模拟座头鲸“气泡网”... 针对老旧支墩坝变形机理复杂、传统预测模型精度不足的问题,提出了基于鲸鱼优化算法(Whale OptimizationAlgorithm,WOA)改进的Transformer变形预测模型。该方法利用WOA对Transformer模型的核心超参数进行全局寻优,模拟座头鲸“气泡网”狩猎策略,通过全局寻优机制优化Transformer核心超参数,结合Transformer自注意力机制对长程依赖的强捕捉能力,提升模型对大坝长期变形序列的拟合与预测性能。以乌溪江支墩坝的实测数据为例进行实证研究,并与GWO-LSTM、标准Transformer等基准模型进行比较。结果显示,所提的WOATransformer模型在各项评价指标上均显著优于对比模型,展现了更强的预测稳定性和泛化能力。该研究成果可为长期服役的大坝工程智能化健康监测提供可靠的技术支持。 展开更多
关键词 支墩坝 大坝变形预测 鲸鱼优化算法 深度学习 Transformer模型 超参数寻优 工程应用
暂未订购 下载PDF
已选0条
引用分析
参考文献
引证文献
 统计分析
检索结果
已选文献
上一页 1 2 … 250 下一页 到第 页
在线咨询 使用帮助 返回顶部 意见反馈