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40,559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：
Three-dimensional cooperative guidance law with practical predefined-time approach angle and time-to-go convergence 认领 引用 被引量：1
1
作者 Zihao WU Wei FENG +3 位作者 Xiaofeng ZHANG Wenxing FU Zhang REN Tuo HAN 《Chinese Journal of Aeronautics》 SCIE EI CAS CSCD 2026年第5期530-546,共17页
This paper presents a Three-Dimensional(3D)cooperative guidance law with Practical Predefined-Time(PPT)convergence for multiple missiles considering approach angle(terminal lineof-sight)and simultaneous arrival constr... This paper presents a Three-Dimensional(3D)cooperative guidance law with Practical Predefined-Time(PPT)convergence for multiple missiles considering approach angle(terminal lineof-sight)and simultaneous arrival constraints.To achieve a salvo attack against a maneuvering target from various directions,the guidance problem is tackled by addressing two critical factors:ensuring that the time-of-arrival is consistent and that the desired approach angles can be met.Considering the short duration of the homing guidance process,the convergence with predefined time for guidance states(especially the approach angle and time-to-go)is factored in.First,for the simultaneous arrival,a PPT guidance law is developed,which can meet the same time-to-go convergence rate in the Line-of-Sight(LOS)direction.Then,in the normal LOS direction,a 3D PPT guidance law is presented considering the approach angle constraint so that the desired approach angles can be reached within a user-designed time.The time-based generator technique is employed in the proposed PPT Cooperative Guidance Law(PPTCGL)to avoid the time-varying gain singularity issue.Notably,this technique can allow the convergence time to be preset in advance,independent of initial system conditions and tuning parameters.Additionally,to avoid excessive gain and improve the robustness of guidance law,a PPT disturbance observer is designed against uncertainties and target maneuvers so that the guidance system perturbation can be compensated in real time.It is userfriendly that the convergence of disturbance estimation can be met with a flexible pre-setting time before achieving the terminal guidance constraints.Finally,extensive numerical simulations are conducted to verify the effectiveness and robustness of the proposed PPTCGL in both the nominal cases and the Monte Carlo test. 展开更多
关键词 Approach angle constraint Cooperative guidance law Maneuvering target Practical predefined-time consensus Three-dimensional guidance model
暂未订购 下载PDF
A Bridge Transformer Network With Deep Graph Convolution for Hyperspectral Image Classification 认领 引用
2
作者 Yuquan Gan Siyu Wu +3 位作者 Chang Su Nan Xiang Zhijie Xu Yushan Pan 《CAAI Transactions on Intelligence Technology》 SCIE EI CSCD 2026年第2期464-482,共19页
Transformers have been widely applied to hyperspectral image classification,leveraging their self-attention mechanism for powerful global modelling.However,two key challenges remain as follows:excessive memory and com... Transformers have been widely applied to hyperspectral image classification,leveraging their self-attention mechanism for powerful global modelling.However,two key challenges remain as follows:excessive memory and computational costs from calculating correlations between all tokens(especially as image size or spectral bands increase)and limited ability to model local boundary information due to lacking explicit enhancement mechanisms.This paper proposes a novel method,bridge transformer network fused with deep graph convolution(BTDGC),to address these issues.The framework includes three components as follows:a double random masking mechanism(DRMM)that forces the model to infer masked features from context during training,a bridge transformer(BT)module with bridge tokens for cross-region feature interaction and a Deep Graph Convolutional Pooling(DGCP)module that preserves spatial topology while aggregating hierarchical information.Experiments on standard hyperspectral datasets show BTDGC outperforms mainstream methods in classification accuracy and robustness,effectively balancing global modelling and local boundary representation.The code is available at http://gffzz188fe103f8f1460asnvfxbkvkp9cc609o.ffgz.tsg.suse.edu.cn/jenny3489/BTDGC. 展开更多
关键词 convolution graph convolutional network masking mechanism transforms
暂未订购 下载PDF
Modularized Graph Convolutional Network 认领 引用
3
作者 Tiantian He Zhixuan Duan Xin Luo 《IEEE/CAA Journal of Automatica Sinica》 SCIE EI CSCD 2026年第3期737-739,共3页
Dear Editor,This letter presents a novel graph neural network, namely modularized graph convolution network(MGCN), to address the underexplored issue in graph convolution networks(GCNs), wherein the weights for neighb... Dear Editor,This letter presents a novel graph neural network, namely modularized graph convolution network(MGCN), to address the underexplored issue in graph convolution networks(GCNs), wherein the weights for neighbor aggregation are fixed, leading to the limited capability of capturing diverse relationships among nodes for representation learning. Conventional GCNs always learn node representations in the graph according to the weights computed from the graph Laplacian, consequently overlooking the similarity and group cohesiveness of node features. 展开更多
关键词 graph convolution networks gcns capturing diverse relationships nodes representation learning modularized graph convolution network neighbor aggregation graph neural network modularized graph convolution network mgcn graph convolutional network
暂未订购 下载PDF
Research on dynamic three-dimensional terrain correction methods of quantitative inversion for airborne gamma-ray spectrometer 认领 引用
4
作者 He-Xi Wu Wei-Cheng Li +6 位作者 Rui Qiu Chao Xiong Yi-Ming Lyu Yi-Qiang Xing De-Hao Zhang Zong-Shuo Tao Yang Wang 《Nuclear Science and Techniques》 SCIE EI CAS CSCD 2026年第5期199-214,共16页
Aerial surveys are dynamic and continuous processes,and there are different height distributions of the ground in the measurement area,which leads to problems such as overlapping measurement areas and inaccurate altit... Aerial surveys are dynamic and continuous processes,and there are different height distributions of the ground in the measurement area,which leads to problems such as overlapping measurement areas and inaccurate altitude correction during the survey process.Commonly used terrain correction methods are based on the concept of finite elementization of ground surface radioactive sources,using GPS coordinates,radar altitude,and ground elevation distribution information from aerial surveys,combined with the sourceless efficiency calibration method to construct a response matrix,which is then inverted for surface nuclide content.However,most of the sourceless efficiency calibration methods used are numerical calculations that consider the body detector as a point detector and do not consider the changes in intrinsic detection efficiency under different incident directions of gamma rays.Therefore,when the altitude of the measurement area varies significantly or the flight altitude of the aerial survey is relatively low,such sourceless efficiency calibration method calculations tend to have a large bias,which affects the accuracy of the terrain correction.To address the above problems,this study employs a novel sourceless efficiency calibration method based on the Boolean operation of the ray deposition process and simplifies the traditional body source measurement model to a surface source measurement model to achieve fast and accurate efficiency calibration.Then,through the discretization of the measurement process,the static measurement process is superposed as equivalent to the dynamic measurement process,and the dynamic measurement response matrix is built and optimized based on the calibration method.Finally,the PSO-MLEM algorithm was used to solve the dynamic measurement response matrix to achieve dynamic terrain correction of aerial survey data.Analysis of the Baiyun'ebo test area revealed that,after applying dynamic terrain correction,the inverted anomalies in uranium(eU),thorium(eTh),and potassium(K)concentrations were closer to ground measurements(within 5.72%-30.79%)and exhibited clearer anomaly boundaries compared to traditional height-based corrections.However,owing to the inherent statistical fluctuations and characteristics of matrix inversion,higher measurement values tend to absorb lower ones,potentially enlarging the anomalous regions.Nevertheless,the highanomaly regions after inversion largely coincided with the ground truth validation,demonstrating that the proposed method can effectively correct airborne gamma spectrometry data. 展开更多
关键词 Airborne gamma-ray spectrum Dynamic three-dimensional Terrain correction
暂未订购 下载PDF
Characteristics of three-dimensional oceanic eddy in the Southern Ocean from 2021 to 2023 认领 引用
5
作者 Tao Jiang Weizeng Shao +3 位作者 Yuyi Hu Xingwei Jiang Qingping Zou Yongliang Wei 《Acta Oceanologica Sinica》 SCIE CAS CSCD 2026年第4期1-16,共16页
This study explores the three-dimensional(3-D)characteristics of oceanic eddies in the Southern Ocean from 2021 to 2023.Copernicus Marine Environment Monitoring Service(CMEMS)GLORYS12V1 product,which provides daily cu... This study explores the three-dimensional(3-D)characteristics of oceanic eddies in the Southern Ocean from 2021 to 2023.Copernicus Marine Environment Monitoring Service(CMEMS)GLORYS12V1 product,which provides daily current field data at a(1/12)°grid resolution,is used to identify eddies with radii>10 km.Additionally,the daily sea level anomaly product from Haiyang-2(HY-2)altimeters is used to detect mesoscale eddies with radii>40 km.GLORYS12V1 detects over ten times more surface eddies than HY-2,likely due to its higher spatial and temporal resolution,which allows better identification of smaller-scale features.Both eddy radius and eddy kinetic energy(EKE)differences between layers decrease with depth.At 0.5 m,EKE is lower than at 300–600 m,where it stabilizes.Over 90%of eddies at these depths show center deflection angles under 3°,defined as the angular offset between eddy centers in adjacent layers relative to the vertical(0°)axis.In a 3-D eddy,the center may shift with depth due to physical processes,causing non-zero center deflection angles between layers.Below 300 m,eddy radius differences are more frequently under 20 km than in the upper 0.5–300 m,where baroclinic instability amplifies,and barotropic instability suppresses cross-layer variability.The influence of both instabilities weakens with depth.In the upper ocean(0.5–300 m),baroclinic instability increases the angular offsets between eddy centers.In contrast,barotropic instability reduces these offsets.At 300–600 m,both promote better vertical alignment,indicating greater structural stability.Overall,this study enhances the understanding of the vertical structure and dynamics of oceanic eddies in the Southern Ocean. 展开更多
关键词 three-dimensional characteristics oceanic eddy Southern Ocean
暂未订购 下载PDF
Three-dimensional full-scale neutronics/thermal-hydraulics/mechanics coupling analysis-based structural assessment of helium-cooled ceramic breeder blanket for fusion reactor 认领 引用
6
作者 Qiang Lian Kui Zhang +6 位作者 Shan-Shan Bu Liang-Ming Pan Wen-Xi Tian Sui-Zheng Qiu Guang-Hui Su Xing-Hua Wu Xiao-Yu Wang 《Nuclear Science and Techniques》 SCIE EI CAS CSCD 2026年第6期178-195,共18页
To accelerate the development and utilization of fusion energy,the China Fusion Engineering Test Reactor(CFETR)has been proposed as a bridge between the International Thermonuclear Experimental Reactor and demonstrati... To accelerate the development and utilization of fusion energy,the China Fusion Engineering Test Reactor(CFETR)has been proposed as a bridge between the International Thermonuclear Experimental Reactor and demonstration fusion reactors.The primary objective of the CFETR is to achieve fusion energy transformation and tritium self-sufficiency,which is realized through the function of the blanket.In this study,a neutronicshermal-hydraulics/mechanics coupling method is developed and applied to a helium-cooled ceramic breeder(HCCB)blanket,which is one of the two blanket candidates for the CFETR.A three-dimensional full-scale model is utilized in the coupling analysis to obtain the distributions of the neutronic,thermal-hydraulic,and mechanical parameters.A structural assessment of the CFETR HCCB blanket is then conducted considering steady-state conditions and two transient scenarios.The results demonstrate that following optimization of the blanket structure,the maximum temperatures of the different components remain below the safety limit of the corresponding materials.The structural assessment indicates that the blanket maintains its structural integrity under steady-state conditions.However,immediately after an in-box loss-of-coolant accident,structural failure owing to stress concentration may occur.Additionally,in the early stage of a loss-of-flow accident,the stress at the joint point between the cooling plate and cap exceeds the allowable stress of the material,potentially leading to structural failure within 17 s if no protective response is implemented.These findings provide comprehensive insights into the performance and safety of the CFETR HCCB blanket design. 展开更多
关键词 Structural assessment Fusion blanket Three-dimensional full-scale model Coupling analysis CFETR
暂未订购 下载PDF
Data-and knowledge-driven three-dimensional geological reconstruction method for tunnel engineering 认领 引用
7
作者 Qiming Li Qian Fang +2 位作者 Jun Wang Gan Wang Peipei Shang 《Journal of Rock Mechanics and Geotechnical Engineering》 SCIE CSCD 2026年第6期4318-4336,共19页
The creation of a three-dimensional(3D)geological model plays a crucial guiding role in engineering.However,in practice,due to the sparsity of boreholes and the invisibility of strata,accurately reconstructing a 3D ge... The creation of a three-dimensional(3D)geological model plays a crucial guiding role in engineering.However,in practice,due to the sparsity of boreholes and the invisibility of strata,accurately reconstructing a 3D geological model has always been a challenging task.In this study,a data-and knowledge-driven 3D geological reconstruction method is proposed,where the Inverse Distance Weighting(IDW)method is integrated with computer vision techniques to improve the accuracy and reliability of geological modeling.The reconstruction of the geological model is realized by the reconstruction of continuous cross-sections in one direction.The reconstruction method integrates two deep learning models:a repair model that learns stratigraphic relationships from borehole data to reconstruct cross-sections,and an interpolation model that predicts intermediate sections by capturing stratigraphic distribution and variation patterns.The comparison with the IDW method and the ordinary kriging method on the virtual data verifies that the proposed method can capture the spatial distribution characteristics of the strata.An engineering example proves that the proposed method can be successfully applied to complex stratum modeling.The proposed method enhances and facilitates intuitive observation of both the reconstructed results and their uncertainties.The proposed method can provide guidance for underground engineering construction sites and contribute to their digital transformation. 展开更多
关键词 Three-dimensional geological modeling Machine learning Borehole data Stratigraphic uncertainty Voxel model
暂未订购 下载PDF
The three-dimensional meshfree numerical manifold method based on parallel computing 认领 引用
8
作者 Keqin Zhang Wei Wu +2 位作者 Danfeng Zhang Yanfei Kang Hehua Zhu 《Acta Mechanica Sinica》 SCIE EI CAS CSCD 2026年第2期360-371,共12页
The numerical manifold method,extensively utilized in numerical computations,faces significant challenges in generating complex manifold elements,particularly for three-dimensional applications.To overcome this challe... The numerical manifold method,extensively utilized in numerical computations,faces significant challenges in generating complex manifold elements,particularly for three-dimensional applications.To overcome this challenge,the meshfree numerical manifold method is developed by integrating the moving least-squares method into the numerical manifold method,effectively bypassing the need for meshing complex geometric objects.However,the implementation of the moving least-squares method introduces computational efficiency issues.To mitigate these,parallel computing methods have been incorporated,resulting in a tenfold increase in the speed of assembling the stiffness matrix with central processing unit parallelism,and a twentyfold increase with graphics processing unit parallelism.The static mechanical system equations for the meshfree numerical manifold method are derived using the Galerkin method.The method’s effectiveness and accuracy are then validated through a series of numerical experiments.The experiments demonstrated that the meshfree numerical manifold method achieves a high precision with minimal nodes and integration points.Additionally,positioning nodes outside the domain significantly improves computational accuracy at the boundaries. 展开更多
关键词 Meshfree numerical manifold method Three-dimensional computation Parallel computation Elastostatics Moving least-squares
暂未订购 下载PDF
Dual Channel Graph Convolutional Networks via Personalized PageRank 认领 引用
9
作者 Longlong Lin Xin Luo 《IEEE/CAA Journal of Automatica Sinica》 SCIE EI CSCD 2026年第1期221-223,共3页
Dear Editor,D2This letter presents a node feature similarity preserving graph convolutional framework P G.Graph neural networks(GNNs)have garnered significant attention for their efficacy in learning graph representat... Dear Editor,D2This letter presents a node feature similarity preserving graph convolutional framework P G.Graph neural networks(GNNs)have garnered significant attention for their efficacy in learning graph representations across diverse real-world applications. 展开更多
关键词 convolutional node feature similarity graph convolutional framework learning graph representations neural networks gnns networks graph personalized
暂未订购 下载PDF
Efficient Video Emotion Recognition via Multi-Scale Region-Aware Convolution and Temporal Interaction Sampling 认领 引用
10
作者 Xiaorui Zhang Chunlin Yuan +1 位作者 Wei Sun Ting Wang 《Computers, Materials & Continua》 SCIE EI 2026年第2期2036-2054,共19页
Video emotion recognition is widely used due to its alignment with the temporal characteristics of human emotional expression,but existingmodels have significant shortcomings.On the one hand,Transformermultihead self-... Video emotion recognition is widely used due to its alignment with the temporal characteristics of human emotional expression,but existingmodels have significant shortcomings.On the one hand,Transformermultihead self-attention modeling of global temporal dependency has problems of high computational overhead and feature similarity.On the other hand,fixed-size convolution kernels are often used,which have weak perception ability for emotional regions of different scales.Therefore,this paper proposes a video emotion recognition model that combines multi-scale region-aware convolution with temporal interactive sampling.In terms of space,multi-branch large-kernel stripe convolution is used to perceive emotional region features at different scales,and attention weights are generated for each scale feature.In terms of time,multi-layer odd-even down-sampling is performed on the time series,and oddeven sub-sequence interaction is performed to solve the problem of feature similarity,while reducing computational costs due to the linear relationship between sampling and convolution overhead.This paper was tested on CMU-MOSI,CMU-MOSEI,and Hume Reaction.The Acc-2 reached 83.4%,85.2%,and 81.2%,respectively.The experimental results show that the model can significantly improve the accuracy of emotion recognition. 展开更多
关键词 Multi-scale region-aware convolution temporal interaction sampling video emotion recognition
暂未订购 下载PDF
Dynamic conditions and processes of three-dimensional accumulation of Sinian-Permian natural gas in Penglai gas area,central Sichuan Basin,SW China 认领 引用
11
作者 LUO Bing ZHANG Benjian +8 位作者 ZHOU Gang WU Luya YAN Wei ZHANG Baoshou ZHANG Xihua ZHONG Yuan MA Kui LUO Xiaorong LI Yishu 《Petroleum Exploration and Development》 SCIE 2026年第2期327-342,共16页
Considering the complexities of gas-water relationships in the gas reservoirs,unclear natural gas distribution and difficult exploration expansion of the Sinian–Permian natural gas in the Penglai gas area of the cent... Considering the complexities of gas-water relationships in the gas reservoirs,unclear natural gas distribution and difficult exploration expansion of the Sinian–Permian natural gas in the Penglai gas area of the central Sichuan Basin,this study investigates the gas source,charging processes and enrichment patterns of gas reservoirs based on reservoir characterization,natural gas geochemical analysis,reservoir testing,well logging-seismic data interpretation,as well as basin modeling and dynamic analysis.The results are obtained in three aspects.First,four sets of highly efficient source rocks are developed beneath the salt of the Triassic Jialingjiang Formation,dominated by the Cambrian source rocks.The reservoirs exhibit strong heterogeneity,with six sets of effective reservoirs being isolated from each other yet dynamically connected.Multi-stage strike-slip fault-related fault-fracture-cavity-unconformity systems constitute the hydrocarbon migration network.Second,overpressure generated by hydrocarbon generation in the Cambrian source rocks drove bidirectional hydrocarbon expulsion from the source kitchen.Multiple sources,including cracked gas from paleo-oil reservoirs and residual hydrocarbons within source rocks,contributed to the hydrocarbon supply.The Sinian–Permian system underwent multiple dynamic hydrocarbon accumulation processes,resulting in the formation of extensive“sweet spots”within multi-layered heterogeneous reservoirs,which were subsequently modified by late-stage gas adjustments to their current form.Third,a three-dimensional accumulation model for deep marine natural gas is established,with multi-source hydrocarbon supply,three-dimensional migration,multi-stage accumulation,dynamic adjustment and lithology-controlled distribution.Large-scale reservoirs within positive structural settings,late-stage structurally stable areas,and slope structures are identified as favorable plays for gas exploration. 展开更多
关键词 Sinian Cambrian Permian natural gas dynamic condition three-dimensional accumulation Penglai gas area Sichuan Basin
暂未订购 下载PDF
Enhanced Image Captioning via Integrated Wavelet Convolution and MobileNet V3 Architecture 认领 引用
12
作者 Mo Hou Bin Xu Wen Shang 《Computers, Materials & Continua》 SCIE EI 2026年第2期897-915,共19页
Image captioning,a pivotal research area at the intersection of image understanding,artificial intelligence,and linguistics,aims to generate natural language descriptions for images.This paper proposes an efficient im... Image captioning,a pivotal research area at the intersection of image understanding,artificial intelligence,and linguistics,aims to generate natural language descriptions for images.This paper proposes an efficient image captioning model named Mob-IMWTC,which integrates improved wavelet convolution(IMWTC)with an enhanced MobileNet V3 architecture.The enhanced MobileNet V3 integrates a transformer encoder as its encoding module and a transformer decoder as its decoding module.This innovative neural network significantly reduces the memory space required and model training time,while maintaining a high level of accuracy in generating image descriptions.IMWTC facilitates large receptive fields without significantly increasing the number of parameters or computational overhead.The improvedMobileNet V3 model has its classifier removed,and simultaneously,it employs IMWTC layers to replace the original convolutional layers.This makes Mob-IMWTC exceptionally well-suited for deployment on lowresource devices.Experimental results,based on objective evaluation metrics such as BLEU,ROUGE,CIDEr,METEOR,and SPICE,demonstrate that Mob-IMWTC outperforms state-of-the-art models,including three CNN architectures(CNN-LSTM,CNN-Att-LSTM,CNN-Tran),two mainstream methods(LCM-Captioner,ClipCap),and our previous work(Mob-Tran).Subjective evaluations further validate the model’s superiority in terms of grammaticality,adequacy,logic,readability,and humanness.Mob-IMWTC offers a lightweight yet effective solution for image captioning,making it suitable for deployment on resource-constrained devices. 展开更多
关键词 Image caption wavelet convolution MobileNet V3 deep learning
暂未订购 下载PDF
Ship Magnetic Field Modeling and Extrapolation Based on a Convolutional Neural Network 认领 引用
13
作者 Ao Zhou Yadong Zhang +3 位作者 Wentie Yang Zuoshuai Wang Jianxun Wang Zhiwei Chen 《哈尔滨工程大学学报（英文版）》 CSCD 2026年第2期536-549,共14页
Accurate modeling of ship magnetic fields is important for predicting their spatial distribution to improve the magnetic stealth effect of ships.This study proposes an extrapolation model for ship magnetic fields base... Accurate modeling of ship magnetic fields is important for predicting their spatial distribution to improve the magnetic stealth effect of ships.This study proposes an extrapolation model for ship magnetic fields based on genetic algorithms and convolutional neural networks(CNNs).The magnetic probe position matrix of the traditional equivalent source is utilized as input,and the three-directional components of the magnetic field measured by the probes are employed as output.The extrapolation model for ship magnetic fields is obtained through iterative training and fitting with CNNs.Variables such as the number of magnetic dipoles,the distance between magnetic dipoles,the size and quantity of convolutional kernels,batch size,learning rate,and L2 regularization coefficient are optimized to boost the accuracy of the extrapolation model for magnetic fields.The fitting accuracy of the extrapolation model for ship magnetic fields is used as the optimization objective.Based on a finite element simulation model of ship magnetic fields,the accuracy and robustness of the CNN algorithm under different magnetic field conditions are validated using the known standard depth plane,the unknown depth at 1.125 times the standard depth plane,and the unknown depth at 1.25 times the standard depth plane.Results show that,after optimization,the fitting error for the magnetic field extrapolation model based on CNN is 1.50%for the standard depth plane,1.63%for the unknown depth at 1.125 times the standard depth plane,and 2.36%for the unknown depth at 1.25 times the standard depth plane.The error remains below 5%under varying magnetic field conditions.When a random measurement error of 0%-5%is introduced for the magnetic probes,the prediction error at 1.25 times the standard depth plane is 2.30%;with a random error of 0%-10%,the prediction error is 4.95%.This approach significantly improves the accuracy and robustness of magnetic field extrapolation,which makes it an effective and feasible method for ship magnetic field modeling. 展开更多
关键词 Shipboard magnetic field Convolutional neural network Genetic algorithm Equivalent source method Magnetic field extrapolation
暂未订购 下载PDF
Spatial-Temporal Graph Fusion with Dual-Scale Convolution for Traffic Flow Prediction 认领 引用
14
作者 Dan Wang Mengyi Cui +1 位作者 Zhenhua Yu Yukang Liu 《Computers, Materials & Continua》 SCIE EI 2026年第6期1375-1396,共22页
Traffic flow prediction is of great importance in traffic planning,road resource management,and congestion mitigation.However,existing prediction have significant limitations in modeling multi-scale spatial-temporal f... Traffic flow prediction is of great importance in traffic planning,road resource management,and congestion mitigation.However,existing prediction have significant limitations in modeling multi-scale spatial-temporal features,particularly in capturing temporal periodicity and spatial dependency in dynamically evolving traffic networks.This paper proposes a novel framework of traffic flow prediction,referred to as Adaptive Graph Fusion Dual-scale Convolutional Network(AGFDCN),which integrates spatial-temporal dynamic graphs with dual-scale convolutional networks.Specifically,we introduce a Dual-Scale Temporal Network,which combines long-and short-term dilated causal convolutions with a temporal decay-aware attention mechanism to efficiently capture traffic patterns across multiple temporal scales.Furthermore,we design a Dynamic Adaptive Graph Module,which models complex spatial dependencies in traffic networks through an adaptive graph fusion mechanism and a dual-path attention-gated module.Finally,the temporal and spatial representations are integrated by employing a gated fusion mechanism,enhancing the overall prediction performance.Experimental results obtained based on three highway datasets(i.e.,PEMS04,PEMS07 and PEMS08)verify that the proposed model outperforms several state-of-the-art baselines in various evaluation metrics.Compared to the spatial-temporal graph model AGCRN with best performance in the baseline models,the proposed model exhibits significant improvements across all datasets:it achieves reduces of MAE by 42.07%and RMSE by 35.43%on PEMS04;MAE by 28.35%and RMSE by 29.28%on PEMS07;and MAE by 30.52%and RMSE by 30.73%on PEMS08,respectively,validating its effectiveness in modeling complex spatial-temporal traffic data and its robustness in handling sudden traffic changes. 展开更多
关键词 Dual-scale convolution dual-path attention-gated module adaptive graph fusion spatial-temporal dynamic graph traffic flow prediction
暂未订购 下载PDF
Prediction of sea surface pCO2in the South China Sea using Spatiotemporal Convolutional LSTM model 认领 引用
15
作者 Shuang LI Yu GAO +4 位作者 Jiannan GAO Yaqi ZHAO Peng HAO Jinbao SONG Chengcheng YU 《Journal of Oceanology and Limnology》 SCIE CAS CSCD 2026年第1期19-35,共17页
The prediction of sea surface partial pressure of carbon dioxide(pCO2)in the South China Sea is crucial for understanding the region’s contribution to the global carbon budget and its interactions with climate cha... The prediction of sea surface partial pressure of carbon dioxide(pCO2)in the South China Sea is crucial for understanding the region’s contribution to the global carbon budget and its interactions with climate change.We applied the Spatiotemporal Convolutional Long Short-Term Memory(STConvLSTM)model,integrating key environmental factors including sea surface temperature(SST),sea surface salinity(SSS),and chlorophyll a(Chl a),to predict and analyze sea surface pCO2in the South China Sea.The model demonstrated high accuracy in short-term predictions(1 month),with a mean absolute error(MAE)of 0.394,a root mean square error(RMSE)of 0.659,and a coefficient of determination(R2)of 0.998.For long-term predictions(12 months),the model maintained its predictive capability,with an MAE of 0.667,RMSE of 1.255,and R2of 0.994.Feature importance analysis revealed that sea surface pCO2and SST were the main drivers of the model’s predictions,whereas Chl a and SSS had relatively minor impacts.The model’s generalization ability was further validated in the northwest Pacific Ocean and tropical Pacific Ocean,where it successfully captured the spatiotemporal variation in pCO2with small prediction errors.The ST-ConvLSTM model provides an efficient and accurate tool for forecasting and analyzing sea surface pCO2in the South China Sea,offering new insights into global carbon cycling and climate change.This study demonstrates the potential of deep learning in marine science and provides a significant technical support for global changes and marine ecosystem research. 展开更多
关键词 sea surface carbon dioxide South China Sea Spatiotemporal Convolutional Long Short-Term Memory(ST-ConvLSTM) deep learning
暂未订购 下载PDF
Multi-Source Traffic Information Completion and Perception Method via Graph Convolutional Neural Networks in Intelligent Connected Transportation System 认领 引用
16
作者 Pangwei Wang Jie Wang +2 位作者 Zipeng Wang Hangrui Dong Li Wang 《Computers, Materials & Continua》 SCIE EI 2026年第8期1417-1435,共19页
Traffic holographic perception refers to the real-time,high-fidelity,and multi-dimensional sensing of traffic states through the fusion of heterogeneous sensors,including cameras,radars,and connected vehicle data.The ... Traffic holographic perception refers to the real-time,high-fidelity,and multi-dimensional sensing of traffic states through the fusion of heterogeneous sensors,including cameras,radars,and connected vehicle data.The multi-source perception data obtained thereby can provide a complete digital representation of the road network for the Intelligent Transportation System(ITS).However,sensors are vulnerable to environmental interference,which can result in data loss at specific points or along arterial highways for certain periods,potentially undermining system safety and decision-making reliability.To address these challenges,a deep learning method based on Graph Convolutional Networks(GCN)and Gated Recurrent Units(GRU)is proposed,leveraging Artificial Intelligence(AI)and intelligent connected technologies for real-time acquisition of multi-sensor perception data.A feature-level fusion integrates multi-source perception data.GCN captures spatial dependencies from the road network topology,while GRU extracts temporal features from time series,enabling accurate imputation of missing traffic data.The method is evaluated at intelligent connected intersections in the Beijing High-level Autonomous Driving Demonstration Area.Results show that the accuracy of long-term traffic state completion reaches 89.36%,and the Root Mean Square Error(RMSE)is reduced by 17.2%compared to the Long Short-Term Memory(LSTM)baseline.This framework provides a practical solution for deploying traffic holographic perception technology in secure and trustworthy ITS. 展开更多
关键词 Intelligent transportation information security traffic information completion traffic holographic perception AI-driven edge computing graph convolutional neural network
暂未订购 下载PDF
Three-dimensional time-dependent fracturing model for hard rock involving stress-induced anisotropic cracks 认领 引用
17
作者 Chen Fan Xia-Ting Feng +2 位作者 Jun Zhao Chengxiang Yang Mengfei Jiang 《Journal of Rock Mechanics and Geotechnical Engineering》 SCIE CSCD 2026年第5期3333-3347,共15页
The time-dependent failure of surrounding rock in deep engineering is essentially controlled by the evolution of microcracks,with the pre-existing fracturing state induced by excavation playing a crucial role in the s... The time-dependent failure of surrounding rock in deep engineering is essentially controlled by the evolution of microcracks,with the pre-existing fracturing state induced by excavation playing a crucial role in the subsequent time-dependent fracturing process.From the perspective of microcrack development,it is a continuous,dynamic process.Therefore,taking the microcrack propagation process as the fundamental principle,this paper proposes a novel three-dimensional(3D)time-dependent model for hard rock that can depict the entire fracturing process within a unified theoretical framework.This developed model discards the traditional tri-modal partition method based on deformation,and instead adopts an analysis approach centred on time-dependent tensile and shear fracturing.The results show that the time-dependent deformation of hard rock is the macroscopic manifestation of the progressive evolution of microcracks over time.Under true triaxial stress,the growth tendency of cracks in hard rock is orientation-dependent throughout the entire loading process.This developed model provides a mechanical explanation for key time-dependent fracture characteristics observed in true triaxial creep tests,including the anisotropy of time-dependent deformation and the preferred orientation of macroscopic failure plane,and provides a novel framework for elucidating the time-dependent failure process of hard rock. 展开更多
关键词 Three-dimensional(3D)time-dependent model Continuous fracturing process Microcrack development Time-dependent anisotropic fracturing True triaxial stress
暂未订购 下载PDF
Mechanisms of efficient three-dimensional fracture network construction in deep shale reservoirs via methane multistage explosive fracturing 认领 引用
18
作者 Yabo Chai Ning Luo +6 位作者 Jianan Zhou Yucheng Wei Hu Zhang Chen Lin Guangrui Ma Cheng Zhai Yu Wang 《Journal of Rock Mechanics and Geotechnical Engineering》 SCIE CSCD 2026年第4期2831-2858,共28页
Methane in situ multistage explosive fracturing(MISMEF)presents a promising technique for enhancing complex fracture networks in deep,low-permeability shale reservoirs.This study employed highfidelity3D simulations,in... Methane in situ multistage explosive fracturing(MISMEF)presents a promising technique for enhancing complex fracture networks in deep,low-permeability shale reservoirs.This study employed highfidelity3D simulations,integrating a characteristic methane–oxygen explosion load model with dynamic relaxation and full-restart methods,to elucidate the coupled interactions between explosive loading and in situ stress.A damage-based zoning approach was developed to quantify fracture characteristics,leading to the proposal of a novel dimensionless evaluation index,Fmef.Results showed that in situ stress predominantly suppressed longitudinal fracture growth,while multistage loading effectively enhanced both lateral and longitudinal propagation following a"delayed initiation–accelerated propagation"pattern.Fracture volume exhibited nonlinear amplificationwith increasing stages,and the continuous increase in fractal dimension suggested improved network connectivity.Energy redistribution driven by the coupled effects of in situ stress and staged loading promoted complex network formation near the wellbore,with MISMEF progressively reducing fracture thresholds through rock mass weakening.Fmef analysis confirmedsignificantimprovement in fracture network quality across all stress conditions,particularly under medium to high in situ stress.This work provides critical mechanistic insights and a theoretical foundation for optimizing MISMEF in deep shale reservoir stimulation. 展开更多
关键词 Methane in situ multistage explosive fracturing(mismef) In situ stress Multistage explosion Three-dimensional fracture network Multidimensional evaluation index
暂未订购 下载PDF
Learning Laws for Deep Convolutional Neural Networks With Guaranteed Convergence 认领 引用
19
作者 Sitan Li Chien Chern Cheah 《IEEE/CAA Journal of Automatica Sinica》 SCIE EI CSCD 2026年第1期170-185,共16页
Convolutional neural networks(CNNs)have shown remarkable success across numerous tasks such as image classification,yet the theoretical understanding of their convergence remains underdeveloped compared to their empir... Convolutional neural networks(CNNs)have shown remarkable success across numerous tasks such as image classification,yet the theoretical understanding of their convergence remains underdeveloped compared to their empirical achievements.In this paper,the first filter learning framework with convergence-guaranteed learning laws for end-to-end learning of deep CNNs is proposed.Novel update laws with convergence analysis are formulated based on the mathematical representation of each layer in convolutional neural networks.The proposed learning laws enable concurrent updates of weights across all layers of the deep convolutional neural network and the analysis shows that the training errors converge to certain bounds which are dependent on the approximation errors.Case studies are conducted on benchmark datasets and the results show that the proposed concurrent filter learning framework guarantees the convergence and offers more consistent and reliable results during training with a trade-off in performance compared to stochastic gradient descent methods.This framework represents a significant step towards enhancing the reliability and effectiveness of deep convolutional neural network by developing a theoretical analysis which allows practical implementation of the learning laws with automatic tuning of the learning rate to guarantee the convergence during training. 展开更多
关键词 Convergence convolution neural networks(CNNs) end-to-end learning online learning
暂未订购 下载PDF
Early gastric dilation after laparoscopic sleeve gastrectomy:Insights from a three-dimensional computed tomography reconstruction study 认领 引用
20
作者 Zhao Li Wen-Zhi Wu +3 位作者 Yi Song Zhao-Peng Li Dong Guo Yu Li 《World Journal of Gastrointestinal Surgery》 SCIE 2026年第1期111-122,共12页
BACKGROUND Laparoscopic sleeve gastrectomy(LSG)is currently the most commonly performed bariatric surgery owing to its effective weight loss and low complication rates.Nonetheless,some patients experience weight regai... BACKGROUND Laparoscopic sleeve gastrectomy(LSG)is currently the most commonly performed bariatric surgery owing to its effective weight loss and low complication rates.Nonetheless,some patients experience weight regain or insufficient weight loss due to residual gastric dilation,the factors of which remain unclear.AIM To evaluate changes in residual gastric volume after LSG using three-dimensional computed tomography reconstruction and to investigate the factors contributing to gastric dilation.METHODS This retrospective study included 50 patients who underwent LSG.Preoperative clinical and laboratory data were obtained.The residual gastric volume was measured using three-dimensional computed tomography reconstruction at 1 month and 3 months postoperatively.The total sleeve volume,tube volume,antral volume,and tube-to-antral volume ratio were also assessed.Resected gastric volume and staple line length were measured during surgery.Weight metrics and laboratory indices were recorded at 1 month,3 months,6 months,and 12 months postoperatively.The Eating Behavior After Bariatric Surgery Questionnaire and Gastroesophageal Reflux Disease Questionnaire(GERD-Q)were used to assess the dietary behavior of patients after LSG.Correlation between the degree of residual gastric dilation and percent total weight loss(%TWL)at 12 months postoperatively was analyzed.Univariate and multivariate correlation analyses were conducted to identify risk factors for residual gastric dilation after LSG.RESULTS The 50 included patients had a mean preoperative body mass index of 42.27±7.19 kg/m2 and average%TWL of 34%±7%at 1 year after LSG.At 1 month after LSG,the mean tube volume,antral volume,and total sleeve volume were 45.93±16.75 mL,115.85±44.92 mL,and 161.77±55.37 mL,respectively.At 3 months after LSG,the residual gastric volume showed statistically significant dilation(average dilation degree:13.50%±17.35%).%TWL at 1 year significantly correlated with residual gastric dilation(P<0.05).Univariate and multivariate linear regression analyses revealed that preoperative type 2 diabetes,residual gastric volume at 1 month after LSG,and GERD-Q scores were independent risk factors influencing the degree of residual gastric dilation.CONCLUSION In conclusion,residual gastric dilation after LSG significantly affected the efficacy of weight loss.Preoperative type 2 diabetes,residual gastric volume at 1 month after LSG,and GERD-Q scores were independent risk factors affecting the degree of residual gastric dilation. 展开更多
关键词 Laparoscopic sleeve gastrectomy Residual gastric dilation Three-dimensional computed tomography reconstruction Weight loss efficacy Residual gastric volume Bariatric surgery
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