Human trajectory prediction is essential and promising in many related applications. This is challenging due to the uncertainty of human behaviors, which can be influenced not only by himself, but also by the surround...Human trajectory prediction is essential and promising in many related applications. This is challenging due to the uncertainty of human behaviors, which can be influenced not only by himself, but also by the surrounding environment. Recent works based on long-short term memory(LSTM) models have brought tremendous improvements on the task of trajectory prediction. However, most of them focus on the spatial influence of humans but ignore the temporal influence. In this paper, we propose a novel spatial-temporal attention(ST-Attention) model,which studies spatial and temporal affinities jointly. Specifically,we introduce an attention mechanism to extract temporal affinity,learning the importance for historical trajectory information at different time instants. To explore spatial affinity, a deep neural network is employed to measure different importance of the neighbors. Experimental results show that our method achieves competitive performance compared with state-of-the-art methods on publicly available datasets.展开更多
Accurate traffic flow prediction has a profound impact on modern traffic management. Traffic flow has complex spatial-temporal correlations and periodicity, which poses difficulties for precise prediction. To address ...Accurate traffic flow prediction has a profound impact on modern traffic management. Traffic flow has complex spatial-temporal correlations and periodicity, which poses difficulties for precise prediction. To address this problem, a Multi-head Self-attention and Spatial-Temporal Graph Convolutional Network (MSSTGCN) for multiscale traffic flow prediction is proposed. Firstly, to capture the hidden traffic periodicity of traffic flow, traffic flow is divided into three kinds of periods, including hourly, daily, and weekly data. Secondly, a graph attention residual layer is constructed to learn the global spatial features across regions. Local spatial-temporal dependence is captured by using a T-GCN module. Thirdly, a transformer layer is introduced to learn the long-term dependence in time. A position embedding mechanism is introduced to label position information for all traffic sequences. Thus, this multi-head self-attention mechanism can recognize the sequence order and allocate weights for different time nodes. Experimental results on four real-world datasets show that the MSSTGCN performs better than the baseline methods and can be successfully adapted to traffic prediction tasks.展开更多
Investigating the transmission of infectious diseases, predicting the epidemic trend of an infectious disease and evaluating the effects of control measures are the continuing subject of public health participators. I...Investigating the transmission of infectious diseases, predicting the epidemic trend of an infectious disease and evaluating the effects of control measures are the continuing subject of public health participators. In recent years the spatial-temporal model has been proposed as a powerful tool to get this information from large incomplete and insufficiently general monitoring data.展开更多
As a significant city in the Yangtze River Delta regions,Hefei has experienced rapid changes in the sources of air pollution due to its high-speed economic development and urban expansion.However,there has been limite...As a significant city in the Yangtze River Delta regions,Hefei has experienced rapid changes in the sources of air pollution due to its high-speed economic development and urban expansion.However,there has been limited research in recent years on the spatial-temporal distribution and emission of its atmospheric pollutants.To address this,this study conducted mobile observations of urban roads using the Mobile-DOAS instrument from June 2021 to May 2022.The monitoring results exhibit a favourable consistent with TROPOMI satellite data and ground monitoring station data.Temporally,there were pronounced seasonal variations in air pollutants.Spatially,high concentration of HCHO and NO_(2)were closely associated with traffic congestion on roadways,while heightened SO_(2)levels were attributed to winter heating and industrial emissions.The study also revealed that with the implementation of road policies,the average vehicle speed increased by 95.4%,while the NO concentration decreased by 54.4%.In the estimation of urban NO_(x)emission flux,it was observed that in temporal terms,compared with inventory data,the emissions calculated viamobile measurements exhibitedmore distinct seasonal patterns,with the highest emission rate of 349 g/sec in winter and the lowest of 142 g/sec in summer.In spatial terms,the significant difference in emissions between the inner and outer ring roads also suggests the presence of the city’s primary NO_(x)emission sources in the area between these two rings.This study offers data support for formulating the next phase of air pollution control measures in urban areas.展开更多
The success of intelligent transportation systems relies heavily on accurate traffic prediction,in which how to model the underlying spatial-temporal information from traffic data has come under the spotlight.Most exi...The success of intelligent transportation systems relies heavily on accurate traffic prediction,in which how to model the underlying spatial-temporal information from traffic data has come under the spotlight.Most existing frameworks typically utilize separate modules for spatial and temporal correlations modeling.However,this stepwise pattern may limit the effectiveness and efficiency in spatial-temporal feature extraction and cause the overlook of important information in some steps.Furthermore,it is lacking sufficient guidance from prior information while modeling based on a given spatial adjacency graph(e.g.,deriving from the geodesic distance or approximate connectivity),and may not reflect the actual interaction between nodes.To overcome those limitations,our paper proposes a spatial-temporal graph synchronous aggregation(STGSA)model to extract the localized and long-term spatial-temporal dependencies simultaneously.Specifically,a tailored graph aggregation method in the vertex domain is designed to extract spatial and temporal features in one graph convolution process.In each STGSA block,we devise a directed temporal correlation graph to represent the localized and long-term dependencies between nodes,and the potential temporal dependence is further fine-tuned by an adaptive weighting operation.Meanwhile,we construct an elaborated spatial adjacency matrix to represent the road sensor graph by considering both physical distance and node similarity in a datadriven manner.Then,inspired by the multi-head attention mechanism which can jointly emphasize information from different r epresentation subspaces,we construct a multi-stream module based on the STGSA blocks to capture global information.It projects the embedding input repeatedly with multiple different channels.Finally,the predicted values are generated by stacking several multi-stream modules.Extensive experiments are constructed on six real-world datasets,and numerical results show that the proposed STGSA model significantly outperforms the benchmarks.展开更多
In the global challenge of Coronavirus disease 2019(COVID-19)pandemic,accurate prediction of daily new cases is crucial for epidemic prevention and socioeconomic planning.In contrast to traditional local,one-dimension...In the global challenge of Coronavirus disease 2019(COVID-19)pandemic,accurate prediction of daily new cases is crucial for epidemic prevention and socioeconomic planning.In contrast to traditional local,one-dimensional time-series data-based infection models,the study introduces an innovative approach by formulating the short-term prediction problem of new cases in a region as multidimensional,gridded time series for both input and prediction targets.A spatial-temporal depth prediction model for COVID-19(ConvLSTM)is presented,and further ConvLSTM by integrating historical meteorological factors(Meteor-ConvLSTM)is refined,considering the influence of meteorological factors on the propagation of COVID-19.The correlation between 10 meteorological factors and the dynamic progression of COVID-19 was evaluated,employing spatial analysis techniques(spatial autocorrelation analysis,trend surface analysis,etc.)to describe the spatial and temporal characteristics of the epidemic.Leveraging the original ConvLSTM,an artificial neural network layer is introduced to learn how meteorological factors impact the infection spread,providing a 5-day forecast at a 0.01°×0.01°pixel resolution.Simulation results using real dataset from the 3.15 outbreak in Shanghai demonstrate the efficacy of Meteor-ConvLSTM,with reduced RMSE of 0.110 and increased R^(2) of 0.125(original ConvLSTM:RMSE=0.702,R^(2)=0.567;Meteor-ConvLSTM:RMSE=0.592,R^(2)=0.692),showcasing its utility for investigating the epidemiological characteristics,transmission dynamics,and epidemic development.展开更多
In the current situation of decelerating economic expansion,examining the digital economy(DE)as a novel economic model is beneficial for the local economy’s sustainable and high-quality development(HQD).We analyzed p...In the current situation of decelerating economic expansion,examining the digital economy(DE)as a novel economic model is beneficial for the local economy’s sustainable and high-quality development(HQD).We analyzed panel data from the Yellow River(YR)region from 2013 to 2021 and discovered notable spatial variances in the composite index and coupling coordination of the two systems.Specifically,the downstream region exhibited the highest coupling coordination,while the upstream region had the lowest.We identified that favorable factors such as economic development,innovation,industrial upgrading,and government intervention can bolster the coupling.Our findings provide a valuable framework for promoting DE and HQD in the YR region.展开更多
Spatial-temporal traffic prediction technology is crucial for network planning,resource allocation optimizing,and user experience improving.With the development of virtual network operators,multi-operator collaboratio...Spatial-temporal traffic prediction technology is crucial for network planning,resource allocation optimizing,and user experience improving.With the development of virtual network operators,multi-operator collaborations,and edge computing,spatial-temporal traffic data has taken on a distributed nature.Consequently,noncentralized spatial-temporal traffic prediction solutions have emerged as a recent research focus.Currently,the majority of research typically adopts federated learning methods to train traffic prediction models distributed on each base station.This method reduces additional burden on communication systems.However,this method has a drawback:it cannot handle irregular traffic data.Due to unstable wireless network environments,device failures,insufficient storage resources,etc.,data missing inevitably occurs during the process of collecting traffic data.This results in the irregular nature of distributed traffic data.Yet,commonly used traffic prediction models such as Recurrent Neural Networks(RNN)and Long Short-Term Memory(LSTM)typically assume that the data is complete and regular.To address the challenge of handling irregular traffic data,this paper transforms irregular traffic prediction into problems of estimating latent variables and generating future traffic.To solve the aforementioned problems,this paper introduces split learning to design a structured distributed learning framework.The framework comprises a Global-level Spatial structure mining Model(GSM)and several Nodelevel Generative Models(NGMs).NGM and GSM represent Seq2Seq models deployed on the base station and graph neural network models deployed on the cloud or central controller.Firstly,the time embedding layer in NGM establishes the mapping relationship between irregular traffic data and regular latent temporal feature variables.Secondly,GSM collects statistical feature parameters of latent temporal feature variables from various nodes and executes graph embedding for spatial-temporal traffic data.Finally,NGM generates future traffic based on latent temporal and spatial feature variables.The introduction of the time attention mechanism enhances the framework’s capability to handle irregular traffic data.Graph attention network introduces spatially correlated base station traffic feature information into local traffic prediction,which compensates for missing information in local irregular traffic data.The proposed framework effectively addresses the distributed prediction issues of irregular traffic data.By testing on real world datasets,the proposed framework improves traffic prediction accuracy by 35%compared to other commonly used distributed traffic prediction methods.展开更多
Accurate control of slab temperature and heating rate is an important significance to improve product performance and reduce carbon emissions for steel rolling reheating furnace(SRRF).Firstly,a spatial temporal distri...Accurate control of slab temperature and heating rate is an important significance to improve product performance and reduce carbon emissions for steel rolling reheating furnace(SRRF).Firstly,a spatial temporal distributed-nonlinear autoregressive with exogenous inputs correlation model(STD-NARXCM)to spatial temporal distributed-autoregressive with exogenous inputs correlation model(STD-ARXCM)in working point is established.Secondly,a new coordinated time-sharing control architecture in different time periods is proposed,which is along the length of the SRRF to improve the control performance.Thirdly,a hybrid control algorithm of expert-fuzzy is proposed to improve the dynamic of the temperature and the heating rate during time period 0 to t_(1).A hybrid control algorithm of expert-fuzzy-PID is proposed to enhance the control accuracy and the heating rate during time period t_(1) to t_(2).A hybrid control algorithm of expert-active disturbance rejection control(ADRC)is proposed to boost the anti-interference and the heating rate during time period t_(2) to t_(3).Finally,the experimental results show that the coordinated time-sharing algorithm can meet the process requirements,the maximum deviation of temperature value is 8-13.5℃.展开更多
The synthesis of high-quality heteroepitaxial diamond films on iridium composite substrates is a critical step toward advancing diamond for electronic and optical applications.Microwave plasma chemical vapor depositio...The synthesis of high-quality heteroepitaxial diamond films on iridium composite substrates is a critical step toward advancing diamond for electronic and optical applications.Microwave plasma chemical vapor deposition,combined with in situ optical emission spectroscopy,enables precise control over growth modes through plasma parameter tuning.In this study,we examine how methane concentration,microwave power,and gas pressure influence plasma species and,consequently,the growth modes of heteroepitaxial diamond by optical emission spectroscopy and scanning electron microscope.At low nucleation densities,increased methane concentrations promote the transition from faceted polyhedral to ballas structures,driven by elevated C_(2) radical concentrations in the plasma.Conversely,at higher nucleation densities,gas pressure,and substrate temperature dominate growth mode determination,leading to diverse morphologies,such as planar,polycrystalline,octahedral,and step-flow growth.These findings elucidate the interplay among plasma species,growth parameters,and growth mode,offering critical insights for optimizing growth conditions and preparing heteroepitaxial diamond films in a specific growth mode.展开更多
Auroral kilometric radiation(AKR),a fundamental plasma emission in Earth's magnetosphere,exhibits three characteristic modes:the right-handed extraordinary(R-X),left-handed ordinary(L-O)and left-handed extraordina...Auroral kilometric radiation(AKR),a fundamental plasma emission in Earth's magnetosphere,exhibits three characteristic modes:the right-handed extraordinary(R-X),left-handed ordinary(L-O)and left-handed extraordinary(L-X)modes.The role of AKR in magnetosphere−ionosphere−atmosphere coupling depends sensitively on its wave mode.While previous studies have primarily focused on the dominant R-X mode,we present the first systematic identification of all three modes using a practical polarization analysis method based on Arase satellite observations.This method employs a spin-axis-relative Ratio:when the satellite's spin axis aligns with the background magnetic field,a positive(negative)Ratio indicates the right-handed(left-handed)polarization,with reversal under anti-parallel conditions.Combined polarization-frequency analysis reveals that R-X,L-O,and L-X modes can exist in both dayside and nightside regions,with power spectral densities up to 10^(-6)mV^(2)m^(-2)Hz^(-1).This study resolves long-standing ambiguities in AKR mode classification and has implications for understanding AKR-induced electron dynamics.展开更多
Using multi-source reanalysis data,this study examines the relationship between the tropical Pacific-Atlantic SST Dipole Mode(TPA-DM)and summer precipitation in North China(NCSP)on the interannual timescale during the...Using multi-source reanalysis data,this study examines the relationship between the tropical Pacific-Atlantic SST Dipole Mode(TPA-DM)and summer precipitation in North China(NCSP)on the interannual timescale during the period of 1979-2022.The results show that the TPA-DM,the dominant pattern of interannual variability in the tropical Pacific and Atlantic regions,exhibits a significant negative correlation with NCSP.The positive phase of TPA-DM induces subsidence over the Maritime Continent through a zonal circulation pattern,which initiates a Pacific-Japan-like wave train along the East Asian coast.The circulation anomalies lead to moisture deficits and convergence subsidence over North China,leading to below-normal rainfall.Further analysis reveals that cooler SST in the Southern Tropical Atlantic facilitates the persistence of the TPA-DM by stimulating the anomalous Walker circulation associated with wind-evaporation-SST-convection feedback.展开更多
Mycorrhizal symbioses are prevalent in terrestrial ecosystems and play essential roles in plant nutrition and health.However,the relative importance of plant evolutionary history,physiology,and eco-geographical factor...Mycorrhizal symbioses are prevalent in terrestrial ecosystems and play essential roles in plant nutrition and health.However,the relative importance of plant evolutionary history,physiology,and eco-geographical factors in shaping mycorrhizal fungal community assembly remains poorly understood.Here,we investigate how plant phylogeny,trophic mode,biogeographic distribution and environmental niche collectively influence the diversity and composition of mycorrhizal fungal communities across the Orchidaceae,spanning broad phylogenetic and ecological scales.By using family-wide orchid-fungal associations and global occurrence data,our analyses showed that the variation in fungal diversity and community structure can be partially explained by orchids’trophic mode,biogeographic distribution and environmental niche,but not by their overall phylogenetic relatedness.Among trophic modes,partially mycoheterotrophic orchids exhibited the highest level of fungal diversity(the lowest level of fungal specificity)in association with a broad range of phylogenetically dispersed fungal partners.Between biogeographical regions,a significantly higher level of fungal specificity was found for orchid species distributed in Australia than those in Eurasia and Africa.Furthermore,multivariate analyses showed that a small portion of the variation in fungal community structure was significantly related to broad climate,soil and vegetation variables,indicating the existence of large-scale habitat filtering on orchid mycorrhizal communities.Altogether,our findings indicate that mycorrhizal communities in the orchid family are likely shaped by multiple,intertwined factors related to orchid ecophysiology and biogeography on a global scale.展开更多
This paper introduces a framework for modeling random fields,with a particular emphasis on analyzing anisotropic spatial variability.It establishes a clear connection between the correlation function and the Kriging v...This paper introduces a framework for modeling random fields,with a particular emphasis on analyzing anisotropic spatial variability.It establishes a clear connection between the correlation function and the Kriging variogram across various anisotropic modes,providing mathematical models to enhance our understanding of random fields.A new anisotropy index,called LSAI,is introduced to quantify anisotropy based on the autocorrelation length and the orientation of the principal axes within the variogram.An LSAI value closer to one indicates a lower degree of anisotropy.The present study examines how the degree of anisotropy varies with different autocorrelation lengths and angles between the principal axes,providing valuable insights into these relationships.To improve the accuracy of parameter probability distribution estimations,this study integrates limited field test data using a Bayesian inference approach.Additionally,the Markov chain Monte Carlo simulation method is employed to develop a conditional random field(CRF)for the deformation modulus.By incorporating data from field bearing plate tests,the posterior variance data for the deformation modulus are derived.This process facilitates the construction of a detailed and reliable CRF for the deformation modulus.展开更多
Climate models are essential for understanding past,present,and future changes in atmospheric circulation,with circulation modes providing key sources of seasonal predictability and prediction uncertainties for both g...Climate models are essential for understanding past,present,and future changes in atmospheric circulation,with circulation modes providing key sources of seasonal predictability and prediction uncertainties for both global and regional climates.This study assesses the performance of models participating in phase 6 of the Coupled Model Intercomparison Project in simulating interannual variability modes of Northern Hemisphere 500-hPa geopotential height during winter and summer,distinguishing predictable(potentially predictable on seasonal or longer timescales)and unpredictable(intraseasonal and essentially unpredictable at long range)components,using reanalysis data and a variance decomposition method.Although most models effectively capture unpredictable modes in reanalysis,their ability to reproduce dominant predictable modes-specifically the Pacific-North American pattern,Arctic Oscillation,and Western Pacific Oscillation in winter,and the East Atlantic and North Atlantic Oscillations in summer-varies notably.An optimal ensemble is identified to distinguish(a)predictable-external modes,dominated by external forcing,and(b)predictable-internal modes,associated with slow internal variability,during the historical period(1950-2014)and the SSP5-8.5 scenario(2036-2100).Under increased radiative forcing,the leading winter/summer predictable-external mode exhibits a more uniform spatial distribution,remarkably larger trend and annual variance,and enhanced height-sea surface temperature(SST)covariance under SSP5-8.5 compared to historical conditions.The dominant winter/summer predictable-internal modes also exhibit increased variance and height-SST covariance under SSP5-8.5,along with localized changes in spatial configuration.Minimal changes are observed in spatial distribution or variance for dominant winter/summer unpredictable modes under SSP5-8.5.This study,from a predictive perspective,deepens our understanding of model uncertainties and projected changes in circulations.展开更多
The subantarctic mode water(SAMW)represents a large water mass in the Southern Ocean.This body of water forms through deep convection(subduction)in winter and contributes to the uptake and storage of anthropogenic hea...The subantarctic mode water(SAMW)represents a large water mass in the Southern Ocean.This body of water forms through deep convection(subduction)in winter and contributes to the uptake and storage of anthropogenic heat.However,its longterm changes in subduction rate and volume in response to shifting climate conditions are unclear.In this study,we investigated the long-term trend of the subduction rate and volume of the South Pacific–SAMW(SPSAMW)using Simple Ocean Data Assimilation outputs during 1980–2017.The results show the overall increasing trend of the subduction rate of the SPSAMW.The increased subduction of the SPSAMW directly contributes to the volume variation in the SPSAMW.The increased subduction in the South Pacific reached(0.28±0.16)Sv-1 per year,which explains nearly 68%of the volume increase in the SPSAMW.This variability in the SPSAMW reflects alterations in the overlying atmosphere.The positive to negative phase change of the Interdecadal Pacific Oscillation(IPO)in 1980–2017 deepened the Amundsen Sea Low(ASL)via atmospheric teleconnections over the South Pacific.Further analysis reveals that the increased westerly winds during the deepening of ASL resulted in more cold water transport from the south,which deepened the winter mixed layer and thus increased subduction and volume within the SPSAMW subduction region.This finding suggests the association of the long-term trends of SPSAMW subduction and volume with the phase change of the IPO.展开更多
The spatial organization of urban-rural systems is fundamentally shaped by the agglomeration and diffusion effects inherent in human-Earth processes,giving rise to distinct gradient-based and hierarchical structures.U...The spatial organization of urban-rural systems is fundamentally shaped by the agglomeration and diffusion effects inherent in human-Earth processes,giving rise to distinct gradient-based and hierarchical structures.Understanding the complexity of these interactions and their multidimensional drivers is essential for deciphering the mechanisms of integrated urban-rural development.Here,we apply a novel hierarchical spatial system framework based on the human-Earth system,combining social network analysis and multi-level modeling,to examine the evolution of the socio-spatial structure in the Beijing-Tianjin-Hebei region from 2000 to 2020.We developed a comprehensive evaluation system spanning economic,social,environmental,and infrastructural dimensions to characterize spatial patterns across multiple network levels,including city clusters,metropolitan areas,municipal-counties,towns,and villages.Our analysis reveals three key findings:First,the density of foundational network connections increased significantly,reflecting a trend toward spatial concentration driven by policy-led regional integration.Second,network structures at the city-cluster and metropolitan scales exhibited a pattern of“initial expansion followed by convergence”,accompanied by notable shifts in their spatial centers of gravity.In parallel,differentiated patterns of agglomeration and expansion were evident in the township-and village-level networks of Baoding,Tangshan,and Handan,while village-level networks in Anxin,Quyang,and other locations demonstrated distinct developmental trends.Third,community structures demonstrated strong functional homophily and interactive cohesion across multiple dimensions,with metropolitan and township communities undergoing restructuring that reflects a reconfiguration of cross-level influence and functional coupling.Spatially,the system manifests as a gradient structure of interwoven point,line,and area networks,establishing a mechanism for functional differentiation and transmission from rural to urban areas.This study provides theoretical foundations and methodological support for understanding the spatial organization logic of integrated urban-rural development,offering practical reference value for advancing regional coordination and rural revitalization in a scientifically informed manner.展开更多
Flexible materials play a crucial role in protecting against behind armour blunt trauma(BABT).However,their compliance complicates the understanding of failure mechanisms and energy absorption.This study used a combin...Flexible materials play a crucial role in protecting against behind armour blunt trauma(BABT).However,their compliance complicates the understanding of failure mechanisms and energy absorption.This study used a combined experimental and numerical approach to investigate the response and failure modes of a flexible ultra-high-molecular-weight polyethylene(UHMWPE)foam protective sandwich structure(UFPSS)under low-velocity impact(LVI).A finite element(FE)model,accounting for nonlinear large deformation and strain-rate-dependent material behavior,was developed for a woven-UFPSS(featuring a plain-woven fabric structure)subjected to a 50 J impact.Experimental and numerical results showed strong agreement in peak force(error<5%),maximum displacement(error<6%),and buffer time(error<8%).The impact's kinetic energy was mainly converted into internal energy of the fabric and foam materials(~50%),viscous dissipation in the foam core(12%-15%),frictional work at the contact interfaces(5%-6%),and work by the pneumatic fixture clamping force(~38%).This study provides the first investigation of the LVI performance of sandwich structures with all soft material layers,offering significant insights for the application of compliant materials in protective fields.展开更多
In dry-coupled ultrasonic thickness measurement,thick rubber layers introduce high-amplitude parasitic echoes that obscure defect signals and degrade thickness accuracy.Existing methods struggle to resolve overlap-pin...In dry-coupled ultrasonic thickness measurement,thick rubber layers introduce high-amplitude parasitic echoes that obscure defect signals and degrade thickness accuracy.Existing methods struggle to resolve overlap-ping echoes under variable coupling conditions and non-stationary noise.This study proposes a novel dual-criterion framework integrating energy contribution and statistical impulsivity metrics to isolate specimen re-flections from coupling-layer interference.By decomposing A-scan signals into Intrinsic Mode Functions(IMFs),the framework employs energy contribution thresholds(>85%)and kurtosis indices(>3)to autonomously select IMFs containing valid specimen echoes.Hybrid time-frequency thresholding further suppresses interference through amplitude filtering and spectral focusing.Experimental results demonstrate the framework’s robustness,achieving 92.3%thickness accuracy for 5 mm steel specimens with 5 mm rubber coupling,outperforming conventional methods by up to 18.7%.The dual-criterion approach reduces operator dependency by 37%and maintainsΔT<0.03 mm under surface roughness up to 6.3μm,offering a practical solution for industrial nondestructive testing with thick dry-coupled interfaces.展开更多
基金supported by the National Key Research and Development Program of China(2018AAA0101005,2018AAA0102404)the Program of the Huawei Technologies Co.Ltd.(FA2018111061SOW12)+1 种基金the National Natural Science Foundation of China(61773054)the Youth Research Fund of the State Key Laboratory of Complex Systems Management and Control(20190213)。
文摘Human trajectory prediction is essential and promising in many related applications. This is challenging due to the uncertainty of human behaviors, which can be influenced not only by himself, but also by the surrounding environment. Recent works based on long-short term memory(LSTM) models have brought tremendous improvements on the task of trajectory prediction. However, most of them focus on the spatial influence of humans but ignore the temporal influence. In this paper, we propose a novel spatial-temporal attention(ST-Attention) model,which studies spatial and temporal affinities jointly. Specifically,we introduce an attention mechanism to extract temporal affinity,learning the importance for historical trajectory information at different time instants. To explore spatial affinity, a deep neural network is employed to measure different importance of the neighbors. Experimental results show that our method achieves competitive performance compared with state-of-the-art methods on publicly available datasets.
基金supported by the National Natural Science Foundation of China(Grant Nos.62472149,62376089,62202147)Hubei Provincial Science and Technology Plan Project(2023BCB04100).
文摘Accurate traffic flow prediction has a profound impact on modern traffic management. Traffic flow has complex spatial-temporal correlations and periodicity, which poses difficulties for precise prediction. To address this problem, a Multi-head Self-attention and Spatial-Temporal Graph Convolutional Network (MSSTGCN) for multiscale traffic flow prediction is proposed. Firstly, to capture the hidden traffic periodicity of traffic flow, traffic flow is divided into three kinds of periods, including hourly, daily, and weekly data. Secondly, a graph attention residual layer is constructed to learn the global spatial features across regions. Local spatial-temporal dependence is captured by using a T-GCN module. Thirdly, a transformer layer is introduced to learn the long-term dependence in time. A position embedding mechanism is introduced to label position information for all traffic sequences. Thus, this multi-head self-attention mechanism can recognize the sequence order and allocate weights for different time nodes. Experimental results on four real-world datasets show that the MSSTGCN performs better than the baseline methods and can be successfully adapted to traffic prediction tasks.
文摘Investigating the transmission of infectious diseases, predicting the epidemic trend of an infectious disease and evaluating the effects of control measures are the continuing subject of public health participators. In recent years the spatial-temporal model has been proposed as a powerful tool to get this information from large incomplete and insufficiently general monitoring data.
基金supported by the National Natural Science Foundation of China(Nos.U19A2044,42105132,42030609,41975037,and 42105133)the National Key Research and Development Program of China(No.2022YFC3703502)+1 种基金the Plan for Anhui Major Provincial Science&Technology Project(No.202203a07020003)Hefei Ecological Environment Bureau Project(No.2020BFFFD01804).
文摘As a significant city in the Yangtze River Delta regions,Hefei has experienced rapid changes in the sources of air pollution due to its high-speed economic development and urban expansion.However,there has been limited research in recent years on the spatial-temporal distribution and emission of its atmospheric pollutants.To address this,this study conducted mobile observations of urban roads using the Mobile-DOAS instrument from June 2021 to May 2022.The monitoring results exhibit a favourable consistent with TROPOMI satellite data and ground monitoring station data.Temporally,there were pronounced seasonal variations in air pollutants.Spatially,high concentration of HCHO and NO_(2)were closely associated with traffic congestion on roadways,while heightened SO_(2)levels were attributed to winter heating and industrial emissions.The study also revealed that with the implementation of road policies,the average vehicle speed increased by 95.4%,while the NO concentration decreased by 54.4%.In the estimation of urban NO_(x)emission flux,it was observed that in temporal terms,compared with inventory data,the emissions calculated viamobile measurements exhibitedmore distinct seasonal patterns,with the highest emission rate of 349 g/sec in winter and the lowest of 142 g/sec in summer.In spatial terms,the significant difference in emissions between the inner and outer ring roads also suggests the presence of the city’s primary NO_(x)emission sources in the area between these two rings.This study offers data support for formulating the next phase of air pollution control measures in urban areas.
基金partially supported by the National Key Research and Development Program of China(2020YFB2104001)。
文摘The success of intelligent transportation systems relies heavily on accurate traffic prediction,in which how to model the underlying spatial-temporal information from traffic data has come under the spotlight.Most existing frameworks typically utilize separate modules for spatial and temporal correlations modeling.However,this stepwise pattern may limit the effectiveness and efficiency in spatial-temporal feature extraction and cause the overlook of important information in some steps.Furthermore,it is lacking sufficient guidance from prior information while modeling based on a given spatial adjacency graph(e.g.,deriving from the geodesic distance or approximate connectivity),and may not reflect the actual interaction between nodes.To overcome those limitations,our paper proposes a spatial-temporal graph synchronous aggregation(STGSA)model to extract the localized and long-term spatial-temporal dependencies simultaneously.Specifically,a tailored graph aggregation method in the vertex domain is designed to extract spatial and temporal features in one graph convolution process.In each STGSA block,we devise a directed temporal correlation graph to represent the localized and long-term dependencies between nodes,and the potential temporal dependence is further fine-tuned by an adaptive weighting operation.Meanwhile,we construct an elaborated spatial adjacency matrix to represent the road sensor graph by considering both physical distance and node similarity in a datadriven manner.Then,inspired by the multi-head attention mechanism which can jointly emphasize information from different r epresentation subspaces,we construct a multi-stream module based on the STGSA blocks to capture global information.It projects the embedding input repeatedly with multiple different channels.Finally,the predicted values are generated by stacking several multi-stream modules.Extensive experiments are constructed on six real-world datasets,and numerical results show that the proposed STGSA model significantly outperforms the benchmarks.
基金the Self-supporting Program of Guangzhou Laboratory(SRPG22-007)the Collaborative Research Project of the National Natural Science Foundation of China(L2224041)+2 种基金and the Chinese Academy of Sciences(XK2022DxC005)Frontier of Interdisciplinary Research on Monitoring and Prediction of Pathogenic Microorganisms in the Atmosphereand the Fundamental Research Funds for the Central Universities(lzujbky-2023-ey10)。
文摘In the global challenge of Coronavirus disease 2019(COVID-19)pandemic,accurate prediction of daily new cases is crucial for epidemic prevention and socioeconomic planning.In contrast to traditional local,one-dimensional time-series data-based infection models,the study introduces an innovative approach by formulating the short-term prediction problem of new cases in a region as multidimensional,gridded time series for both input and prediction targets.A spatial-temporal depth prediction model for COVID-19(ConvLSTM)is presented,and further ConvLSTM by integrating historical meteorological factors(Meteor-ConvLSTM)is refined,considering the influence of meteorological factors on the propagation of COVID-19.The correlation between 10 meteorological factors and the dynamic progression of COVID-19 was evaluated,employing spatial analysis techniques(spatial autocorrelation analysis,trend surface analysis,etc.)to describe the spatial and temporal characteristics of the epidemic.Leveraging the original ConvLSTM,an artificial neural network layer is introduced to learn how meteorological factors impact the infection spread,providing a 5-day forecast at a 0.01°×0.01°pixel resolution.Simulation results using real dataset from the 3.15 outbreak in Shanghai demonstrate the efficacy of Meteor-ConvLSTM,with reduced RMSE of 0.110 and increased R^(2) of 0.125(original ConvLSTM:RMSE=0.702,R^(2)=0.567;Meteor-ConvLSTM:RMSE=0.592,R^(2)=0.692),showcasing its utility for investigating the epidemiological characteristics,transmission dynamics,and epidemic development.
基金supported by the National Office for Philosophy and Social Sciences(grant reference 22&ZD067).
文摘In the current situation of decelerating economic expansion,examining the digital economy(DE)as a novel economic model is beneficial for the local economy’s sustainable and high-quality development(HQD).We analyzed panel data from the Yellow River(YR)region from 2013 to 2021 and discovered notable spatial variances in the composite index and coupling coordination of the two systems.Specifically,the downstream region exhibited the highest coupling coordination,while the upstream region had the lowest.We identified that favorable factors such as economic development,innovation,industrial upgrading,and government intervention can bolster the coupling.Our findings provide a valuable framework for promoting DE and HQD in the YR region.
基金supported by the Beijing Natural Science Foundation(Certificate Number:L234025).
文摘Spatial-temporal traffic prediction technology is crucial for network planning,resource allocation optimizing,and user experience improving.With the development of virtual network operators,multi-operator collaborations,and edge computing,spatial-temporal traffic data has taken on a distributed nature.Consequently,noncentralized spatial-temporal traffic prediction solutions have emerged as a recent research focus.Currently,the majority of research typically adopts federated learning methods to train traffic prediction models distributed on each base station.This method reduces additional burden on communication systems.However,this method has a drawback:it cannot handle irregular traffic data.Due to unstable wireless network environments,device failures,insufficient storage resources,etc.,data missing inevitably occurs during the process of collecting traffic data.This results in the irregular nature of distributed traffic data.Yet,commonly used traffic prediction models such as Recurrent Neural Networks(RNN)and Long Short-Term Memory(LSTM)typically assume that the data is complete and regular.To address the challenge of handling irregular traffic data,this paper transforms irregular traffic prediction into problems of estimating latent variables and generating future traffic.To solve the aforementioned problems,this paper introduces split learning to design a structured distributed learning framework.The framework comprises a Global-level Spatial structure mining Model(GSM)and several Nodelevel Generative Models(NGMs).NGM and GSM represent Seq2Seq models deployed on the base station and graph neural network models deployed on the cloud or central controller.Firstly,the time embedding layer in NGM establishes the mapping relationship between irregular traffic data and regular latent temporal feature variables.Secondly,GSM collects statistical feature parameters of latent temporal feature variables from various nodes and executes graph embedding for spatial-temporal traffic data.Finally,NGM generates future traffic based on latent temporal and spatial feature variables.The introduction of the time attention mechanism enhances the framework’s capability to handle irregular traffic data.Graph attention network introduces spatially correlated base station traffic feature information into local traffic prediction,which compensates for missing information in local irregular traffic data.The proposed framework effectively addresses the distributed prediction issues of irregular traffic data.By testing on real world datasets,the proposed framework improves traffic prediction accuracy by 35%compared to other commonly used distributed traffic prediction methods.
基金This work was supported by the National Natural Science Foundation of China(Nos.62173032 and 62003038).
文摘Accurate control of slab temperature and heating rate is an important significance to improve product performance and reduce carbon emissions for steel rolling reheating furnace(SRRF).Firstly,a spatial temporal distributed-nonlinear autoregressive with exogenous inputs correlation model(STD-NARXCM)to spatial temporal distributed-autoregressive with exogenous inputs correlation model(STD-ARXCM)in working point is established.Secondly,a new coordinated time-sharing control architecture in different time periods is proposed,which is along the length of the SRRF to improve the control performance.Thirdly,a hybrid control algorithm of expert-fuzzy is proposed to improve the dynamic of the temperature and the heating rate during time period 0 to t_(1).A hybrid control algorithm of expert-fuzzy-PID is proposed to enhance the control accuracy and the heating rate during time period t_(1) to t_(2).A hybrid control algorithm of expert-active disturbance rejection control(ADRC)is proposed to boost the anti-interference and the heating rate during time period t_(2) to t_(3).Finally,the experimental results show that the coordinated time-sharing algorithm can meet the process requirements,the maximum deviation of temperature value is 8-13.5℃.
基金funded by the National Key Research and Development Program of China(Grant No.2022YFB3608602)the National Natural Science Foundation of China(Grant Nos.62404215 and 62574199)Instrument and Equipment Development Project of CAS(Grant No.PTYQ2024TD0003)。
文摘The synthesis of high-quality heteroepitaxial diamond films on iridium composite substrates is a critical step toward advancing diamond for electronic and optical applications.Microwave plasma chemical vapor deposition,combined with in situ optical emission spectroscopy,enables precise control over growth modes through plasma parameter tuning.In this study,we examine how methane concentration,microwave power,and gas pressure influence plasma species and,consequently,the growth modes of heteroepitaxial diamond by optical emission spectroscopy and scanning electron microscope.At low nucleation densities,increased methane concentrations promote the transition from faceted polyhedral to ballas structures,driven by elevated C_(2) radical concentrations in the plasma.Conversely,at higher nucleation densities,gas pressure,and substrate temperature dominate growth mode determination,leading to diverse morphologies,such as planar,polycrystalline,octahedral,and step-flow growth.These findings elucidate the interplay among plasma species,growth parameters,and growth mode,offering critical insights for optimizing growth conditions and preparing heteroepitaxial diamond films in a specific growth mode.
基金supported by the National Natural Science Foundation of China(Grants 42374215,42230209,42374199,42304183,42422406,42174185,72061147004 and 72342001)the Science and Technology Development Fund,Macao SAR(File no.0042/2024/RIA1 and 0008/2024/AKP)+1 种基金the Natural Science Foundation of Hunan Province(Grant 2023JJ20038)the Research Project of Science and Technology of Hunan Province(2025JJ10009,2022RC4025,2025QK1004,2023JJ50312,2023JJ50010 and 2024RC9012).
文摘Auroral kilometric radiation(AKR),a fundamental plasma emission in Earth's magnetosphere,exhibits three characteristic modes:the right-handed extraordinary(R-X),left-handed ordinary(L-O)and left-handed extraordinary(L-X)modes.The role of AKR in magnetosphere−ionosphere−atmosphere coupling depends sensitively on its wave mode.While previous studies have primarily focused on the dominant R-X mode,we present the first systematic identification of all three modes using a practical polarization analysis method based on Arase satellite observations.This method employs a spin-axis-relative Ratio:when the satellite's spin axis aligns with the background magnetic field,a positive(negative)Ratio indicates the right-handed(left-handed)polarization,with reversal under anti-parallel conditions.Combined polarization-frequency analysis reveals that R-X,L-O,and L-X modes can exist in both dayside and nightside regions,with power spectral densities up to 10^(-6)mV^(2)m^(-2)Hz^(-1).This study resolves long-standing ambiguities in AKR mode classification and has implications for understanding AKR-induced electron dynamics.
基金jointly supported by the Second Tibetan Plateau Scientific Expedition and Research Program[grant number-ber 2019QZKK0103]the National Natural Science Foundation of China[grant number 42293294]the China Meteorological Admin-istration Climate Change Special Program[grant number QBZ202303]。
文摘Using multi-source reanalysis data,this study examines the relationship between the tropical Pacific-Atlantic SST Dipole Mode(TPA-DM)and summer precipitation in North China(NCSP)on the interannual timescale during the period of 1979-2022.The results show that the TPA-DM,the dominant pattern of interannual variability in the tropical Pacific and Atlantic regions,exhibits a significant negative correlation with NCSP.The positive phase of TPA-DM induces subsidence over the Maritime Continent through a zonal circulation pattern,which initiates a Pacific-Japan-like wave train along the East Asian coast.The circulation anomalies lead to moisture deficits and convergence subsidence over North China,leading to below-normal rainfall.Further analysis reveals that cooler SST in the Southern Tropical Atlantic facilitates the persistence of the TPA-DM by stimulating the anomalous Walker circulation associated with wind-evaporation-SST-convection feedback.
基金the funding provided by the China Scholarship Council(Grant No.201804910634)the Ecology Fund of the Royal Netherlands Academy of Arts and Sciences(KNAWWF/807/19039)to Deyi Wang.
文摘Mycorrhizal symbioses are prevalent in terrestrial ecosystems and play essential roles in plant nutrition and health.However,the relative importance of plant evolutionary history,physiology,and eco-geographical factors in shaping mycorrhizal fungal community assembly remains poorly understood.Here,we investigate how plant phylogeny,trophic mode,biogeographic distribution and environmental niche collectively influence the diversity and composition of mycorrhizal fungal communities across the Orchidaceae,spanning broad phylogenetic and ecological scales.By using family-wide orchid-fungal associations and global occurrence data,our analyses showed that the variation in fungal diversity and community structure can be partially explained by orchids’trophic mode,biogeographic distribution and environmental niche,but not by their overall phylogenetic relatedness.Among trophic modes,partially mycoheterotrophic orchids exhibited the highest level of fungal diversity(the lowest level of fungal specificity)in association with a broad range of phylogenetically dispersed fungal partners.Between biogeographical regions,a significantly higher level of fungal specificity was found for orchid species distributed in Australia than those in Eurasia and Africa.Furthermore,multivariate analyses showed that a small portion of the variation in fungal community structure was significantly related to broad climate,soil and vegetation variables,indicating the existence of large-scale habitat filtering on orchid mycorrhizal communities.Altogether,our findings indicate that mycorrhizal communities in the orchid family are likely shaped by multiple,intertwined factors related to orchid ecophysiology and biogeography on a global scale.
基金supported by the Doctoral Research Funds for Nanchang HangKong University,China(Grant No.EA202411211)support is gratefully acknowledged.
文摘This paper introduces a framework for modeling random fields,with a particular emphasis on analyzing anisotropic spatial variability.It establishes a clear connection between the correlation function and the Kriging variogram across various anisotropic modes,providing mathematical models to enhance our understanding of random fields.A new anisotropy index,called LSAI,is introduced to quantify anisotropy based on the autocorrelation length and the orientation of the principal axes within the variogram.An LSAI value closer to one indicates a lower degree of anisotropy.The present study examines how the degree of anisotropy varies with different autocorrelation lengths and angles between the principal axes,providing valuable insights into these relationships.To improve the accuracy of parameter probability distribution estimations,this study integrates limited field test data using a Bayesian inference approach.Additionally,the Markov chain Monte Carlo simulation method is employed to develop a conditional random field(CRF)for the deformation modulus.By incorporating data from field bearing plate tests,the posterior variance data for the deformation modulus are derived.This process facilitates the construction of a detailed and reliable CRF for the deformation modulus.
基金supported by the National Natural Science Foundation of China(Grant Nos.U2342210 and 42275043)the National Institute of Natural Hazards,Ministry of Emergency Management of China(Grant Nos.J2223806,ZDJ2024-25 and ZDJ2025-34)。
文摘Climate models are essential for understanding past,present,and future changes in atmospheric circulation,with circulation modes providing key sources of seasonal predictability and prediction uncertainties for both global and regional climates.This study assesses the performance of models participating in phase 6 of the Coupled Model Intercomparison Project in simulating interannual variability modes of Northern Hemisphere 500-hPa geopotential height during winter and summer,distinguishing predictable(potentially predictable on seasonal or longer timescales)and unpredictable(intraseasonal and essentially unpredictable at long range)components,using reanalysis data and a variance decomposition method.Although most models effectively capture unpredictable modes in reanalysis,their ability to reproduce dominant predictable modes-specifically the Pacific-North American pattern,Arctic Oscillation,and Western Pacific Oscillation in winter,and the East Atlantic and North Atlantic Oscillations in summer-varies notably.An optimal ensemble is identified to distinguish(a)predictable-external modes,dominated by external forcing,and(b)predictable-internal modes,associated with slow internal variability,during the historical period(1950-2014)and the SSP5-8.5 scenario(2036-2100).Under increased radiative forcing,the leading winter/summer predictable-external mode exhibits a more uniform spatial distribution,remarkably larger trend and annual variance,and enhanced height-sea surface temperature(SST)covariance under SSP5-8.5 compared to historical conditions.The dominant winter/summer predictable-internal modes also exhibit increased variance and height-SST covariance under SSP5-8.5,along with localized changes in spatial configuration.Minimal changes are observed in spatial distribution or variance for dominant winter/summer unpredictable modes under SSP5-8.5.This study,from a predictive perspective,deepens our understanding of model uncertainties and projected changes in circulations.
基金supported by the National Natural Science Foundation of China(Nos.42406256,42376034,and 42430402)the Qingdao Postdoctoral Application Research Project(No.QDBSH20220202152)+1 种基金the National Key R&D Program of China(No.2018YFA0605701)the Chinese Arctic and Antarctic Administration(No.IRASCC2020-2022-02-01-03)。
文摘The subantarctic mode water(SAMW)represents a large water mass in the Southern Ocean.This body of water forms through deep convection(subduction)in winter and contributes to the uptake and storage of anthropogenic heat.However,its longterm changes in subduction rate and volume in response to shifting climate conditions are unclear.In this study,we investigated the long-term trend of the subduction rate and volume of the South Pacific–SAMW(SPSAMW)using Simple Ocean Data Assimilation outputs during 1980–2017.The results show the overall increasing trend of the subduction rate of the SPSAMW.The increased subduction of the SPSAMW directly contributes to the volume variation in the SPSAMW.The increased subduction in the South Pacific reached(0.28±0.16)Sv-1 per year,which explains nearly 68%of the volume increase in the SPSAMW.This variability in the SPSAMW reflects alterations in the overlying atmosphere.The positive to negative phase change of the Interdecadal Pacific Oscillation(IPO)in 1980–2017 deepened the Amundsen Sea Low(ASL)via atmospheric teleconnections over the South Pacific.Further analysis reveals that the increased westerly winds during the deepening of ASL resulted in more cold water transport from the south,which deepened the winter mixed layer and thus increased subduction and volume within the SPSAMW subduction region.This finding suggests the association of the long-term trends of SPSAMW subduction and volume with the phase change of the IPO.
基金supported by the National Natural Science Foundation of China(Grant Nos.42293270,42530712)the Young Scientists Fund of the National Natural Science Foundation of China(Grant No.42401334).
文摘The spatial organization of urban-rural systems is fundamentally shaped by the agglomeration and diffusion effects inherent in human-Earth processes,giving rise to distinct gradient-based and hierarchical structures.Understanding the complexity of these interactions and their multidimensional drivers is essential for deciphering the mechanisms of integrated urban-rural development.Here,we apply a novel hierarchical spatial system framework based on the human-Earth system,combining social network analysis and multi-level modeling,to examine the evolution of the socio-spatial structure in the Beijing-Tianjin-Hebei region from 2000 to 2020.We developed a comprehensive evaluation system spanning economic,social,environmental,and infrastructural dimensions to characterize spatial patterns across multiple network levels,including city clusters,metropolitan areas,municipal-counties,towns,and villages.Our analysis reveals three key findings:First,the density of foundational network connections increased significantly,reflecting a trend toward spatial concentration driven by policy-led regional integration.Second,network structures at the city-cluster and metropolitan scales exhibited a pattern of“initial expansion followed by convergence”,accompanied by notable shifts in their spatial centers of gravity.In parallel,differentiated patterns of agglomeration and expansion were evident in the township-and village-level networks of Baoding,Tangshan,and Handan,while village-level networks in Anxin,Quyang,and other locations demonstrated distinct developmental trends.Third,community structures demonstrated strong functional homophily and interactive cohesion across multiple dimensions,with metropolitan and township communities undergoing restructuring that reflects a reconfiguration of cross-level influence and functional coupling.Spatially,the system manifests as a gradient structure of interwoven point,line,and area networks,establishing a mechanism for functional differentiation and transmission from rural to urban areas.This study provides theoretical foundations and methodological support for understanding the spatial organization logic of integrated urban-rural development,offering practical reference value for advancing regional coordination and rural revitalization in a scientifically informed manner.
基金supported by the Zhenjiang Key R&D Plan(GY2021009)Lianyungang City Major Technology Breakthrough(CGJBGS2104)+2 种基金National Natural Science Foundation of China under Grant(12302456)National Key Laboratory Foundation of Science and Technology on Materials under Shock and Impact under Grant(6142902241601)China Postdoctoral Science Foundation under Grants(2025M774217)。
文摘Flexible materials play a crucial role in protecting against behind armour blunt trauma(BABT).However,their compliance complicates the understanding of failure mechanisms and energy absorption.This study used a combined experimental and numerical approach to investigate the response and failure modes of a flexible ultra-high-molecular-weight polyethylene(UHMWPE)foam protective sandwich structure(UFPSS)under low-velocity impact(LVI).A finite element(FE)model,accounting for nonlinear large deformation and strain-rate-dependent material behavior,was developed for a woven-UFPSS(featuring a plain-woven fabric structure)subjected to a 50 J impact.Experimental and numerical results showed strong agreement in peak force(error<5%),maximum displacement(error<6%),and buffer time(error<8%).The impact's kinetic energy was mainly converted into internal energy of the fabric and foam materials(~50%),viscous dissipation in the foam core(12%-15%),frictional work at the contact interfaces(5%-6%),and work by the pneumatic fixture clamping force(~38%).This study provides the first investigation of the LVI performance of sandwich structures with all soft material layers,offering significant insights for the application of compliant materials in protective fields.
基金funded by the National Natural Science Foundation of China,grant number U24A20135Inner Mongolia Natural Science Foundation major project,grant number 2023ZD12+7 种基金Inner Mongolia Autonomous Region key research and development and achievement transformation plan project,grant number 2023YFHH0090Natural Science Foundation of Inner Mongolia,grant number 2022MS05006Inner Mongolia Autonomous Region Talent Development FundUniversity basic research business expenses,grant number 2023RCTD012University basic research business expenses,grant number 2023QNJS075Postgraduate Research Innovation Program and of Inner Mongolia Autonomous Region,grant number KC2024053BUniversity basic research business expenses,grant number 2024YXXS012National Key Laboratory of Special Vehicle Design and Manufacturing Integration Technology,grant number GZ2023KF012.
文摘In dry-coupled ultrasonic thickness measurement,thick rubber layers introduce high-amplitude parasitic echoes that obscure defect signals and degrade thickness accuracy.Existing methods struggle to resolve overlap-ping echoes under variable coupling conditions and non-stationary noise.This study proposes a novel dual-criterion framework integrating energy contribution and statistical impulsivity metrics to isolate specimen re-flections from coupling-layer interference.By decomposing A-scan signals into Intrinsic Mode Functions(IMFs),the framework employs energy contribution thresholds(>85%)and kurtosis indices(>3)to autonomously select IMFs containing valid specimen echoes.Hybrid time-frequency thresholding further suppresses interference through amplitude filtering and spectral focusing.Experimental results demonstrate the framework’s robustness,achieving 92.3%thickness accuracy for 5 mm steel specimens with 5 mm rubber coupling,outperforming conventional methods by up to 18.7%.The dual-criterion approach reduces operator dependency by 37%and maintainsΔT<0.03 mm under surface roughness up to 6.3μm,offering a practical solution for industrial nondestructive testing with thick dry-coupled interfaces.
