The rapid melting of Arctic sea ice poses significant risks to the safety of shipping routes.Accurate remote sensing data on sea ice concentration(SIC)is crucial for effective route planning of ships and ensuring navi...The rapid melting of Arctic sea ice poses significant risks to the safety of shipping routes.Accurate remote sensing data on sea ice concentration(SIC)is crucial for effective route planning of ships and ensuring navigational safety.Despite the availability of numerous SIC products in China,these datasets still lag behind mainstream international products in terms of data accuracy,spatiotemporal resolution,and time span.To enhance the accuracy of China's domestic SIC remote sensing data,this study used the SIC data derived from the passive microwave remote sensing dataset provided by the University of Bremen(BRM-SIC)as a reference to conduct a comprehensive evaluation and analysis of two additional SIC datasets:the dataset derived from the microwave radiation imager(MWRI)aboard the FY-3D satellite,provided by the National Satellite Meteorological Center(FY-SIC),and the dataset obtained through the DT-ASI algorithm from the microwave imager of the FY-3D satellite,provided by Ocean University of China(OUC-SIC).Based on the evaluation results,a TransUnet fusion correction model was developed.The performance of this model was then compared against Ordinary Least Squares(OLS),Random Forest(RF),and UNet correction models,through spatial and temporal analyses.Results indicate that,compared to FY-SIC data,the RMSE of the OUC-SIC data and the standard data is reduced by24.245%,while the R is increased by 12.516%.Overall,the accuracy of OUC-SIC data is superior to that of FY-SIC data.During the research period(2020–2022),the standard deviation(SD)and coefficient of variation(CV)of OUC-SIC were 3.877%and 10.582%,respectively,while those for FY-SIC were 7.836%and 7.982%,respectively.In the study area,compared with OUC-SIC data,FYSIC data exhibited a larger standard deviation of deviation and a smaller coefficient of variation of deviation across most sea areas.These results indicate that the OUC-SIC data exhibit better temporal and spatial stability,whereas the FY-SIC data show stronger relative dimensionless stability.Among the four correction models,all showed improvements over the original,unfused corrected data.The fusion corrections using the OLS,RF,UNet,and TransUnet models reduced RMSE by 5.563%,14.601%,42.927%,and48.316%,respectively.Correspondingly,R increased by 0.463%,1.176%,3.951%,and 4.342%,respectively.Among these models,TransUnet performed the best,effectively integrating the advantages of FY-SIC and OUC-SIC data and notably improving the overall accuracy and spatiotemporal stability of SIC data.展开更多
Tropospheric zenith wet delay(ZWD)plays a vital role in the analysis of space geodetic observations.In recent years,machine learning methods have been increasingly applied to improve the accuracy of ZWD calculations.H...Tropospheric zenith wet delay(ZWD)plays a vital role in the analysis of space geodetic observations.In recent years,machine learning methods have been increasingly applied to improve the accuracy of ZWD calculations.However,a single machine learning model has limited generalization capabilities.To address these limitations,this study introduces a novel machine learning fusion(MLF)algorithm with stronger generalization capabilities to enhance ZWD modeling and prediction accuracy.The MLF algorithm utilizes a two-layer structure integrating extra trees(ET),backpropagation neural network(BPNN),and linear regression models.By comparing the root mean square error(RMSE)of these models,we found that both ET-based and MLF-based models outperform RF-based and BPNN-based models in terms of internal and external accuracy,across both surface meteorological data-based and blind models.The improvement in exte rnal accuracy is particularly significant in the blind models.Our re sults show that the MLF(with an RMSE of 3.93 cm)and ET(3.99 cm)models outperform the traditional GPT3model(4.07 cm),while the RF(4.21 cm)and BPNN(4.14 cm)have worse external accuracies than the GPT3 model.It is worth noting that the BPNN suffered from overfitting during external accuracy tests,which was avoided by the MLF.In summary,regardless of the availability of surface meteorological data,the MLF-based empirical models demonstrate superior internal and external accuracy compared to the other tested models in this study.展开更多
Solar flare prediction is an important subject in the field of space weather.Deep learning technology has greatly promoted the development of this subject.In this study,we propose a novel solar flare forecasting model...Solar flare prediction is an important subject in the field of space weather.Deep learning technology has greatly promoted the development of this subject.In this study,we propose a novel solar flare forecasting model integrating Deep Residual Network(ResNet)and Support Vector Machine(SVM)for both≥C-class(C,M,and X classes)and≥M-class(M and X classes)flares.We collected samples of magnetograms from May 1,2010 to September 13,2018 from Space-weather Helioseismic and Magnetic Imager(HMI)Active Region Patches and then used a cross-validation method to obtain seven independent data sets.We then utilized five metrics to evaluate our fusion model,based on intermediate-output extracted by ResNet and SVM using the Gaussian kernel function.Our results show that the primary metric true skill statistics(TSS)achieves a value of 0.708±0.027 for≥C-class prediction,and of 0.758±0.042 for≥M-class prediction;these values indicate that our approach performs significantly better than those of previous studies.The metrics of our fusion model’s performance on the seven datasets indicate that the model is quite stable and robust,suggesting that fusion models that integrate an excellent baseline network with SVM can achieve improved performance in solar flare prediction.Besides,we also discuss the performance impact of architectural innovation in our fusion model.展开更多
This study presents new methods to effectively model the anisotropic yielding and hardening behavior of laser powder bed fusion fabricated aluminum alloy under both monotonic and cyclic loading conditions.The proposed...This study presents new methods to effectively model the anisotropic yielding and hardening behavior of laser powder bed fusion fabricated aluminum alloy under both monotonic and cyclic loading conditions.The proposed model combines the yield surface-interpolation method to accurately describe the anisotropic hardening rates in various directions,with the Chaboche kinematic hardening rule to precisely reflect the cyclic characteristics.For numerical implementation of the combined anisotropic and cyclic constitutive model,a fully implicit stress integration algorithm based on return mapping method is provided.Moreover,the multiple parameters associated with the model are categorized and identified in an uncoupled manner.The isotropic and cyclic hardening parameters are determined by an inverse method,and the stability of the optimization outcomes is validated by applying different starting points for the parameters.Particularly,the back-stress effect on the identification of anisotropic parameters associated with the stress invariant-based Hill48 yield function is considered for the first time.This consideration leads to an improved prediction accuracy compared to the identification of anisotropic parameters without considering back-stress effect.The combined anisotropic and cyclic constitutive model,along with the calibrated parameters,are proven capable of accurately reproducing the intricate deformation behavior of laser powder bed fusion fabricated AlSi10Mg.展开更多
Semi-crystalline polymer laser powder bed fusion(L-PBF)has recently attracted increasing interest due to its potential for fabricating complex geometry.However,a more comprehensive understanding of the underlying phys...Semi-crystalline polymer laser powder bed fusion(L-PBF)has recently attracted increasing interest due to its potential for fabricating complex geometry.However,a more comprehensive understanding of the underlying physics during L-PBF is required to better control the properties of the final part.This work proposed a multi-layer numerical model to study the temperature and phase evolution during the polyamide-12(PA12)L-PBF process.The Descend and Parallel Chord methods were introduced to improve the convergence of the non-linear thermal solver.The level-set-based mesh adaptation strategy,governed by multi-physical fields,was applied to alleviate the calculation and accurately track the phase evolution.The processing simulation on the dog-bone model revealed that preheating temperature significantly influences the crystallization behavior.Finally,the multi-layer simulation demonstrated that such a developed numerical model can be used to study the phase transformation during powder layer updating and the cyclic laser sintering phenomena.Moreover,the numerical study suggested that crystallization occurs slowly during the L-PBF process.展开更多
With the advancement of scientific research and the rapid growth of the internet,academic users increasingly face challenges in obtaining accurate information about peer research.As a key component of big data analyti...With the advancement of scientific research and the rapid growth of the internet,academic users increasingly face challenges in obtaining accurate information about peer research.As a key component of big data analytics,user profiling has emerged as a critical focus in the scientific research community.While graph neural networks(GNNs)perform well in various graph learning tasks,their scalability to large graphs becomes problematic as the number of nodes increases due to computational complexity.To address this issue,this study proposes a novel academic user profiling model based on graph neural networks tailored to the unique characteristics of scientific research networks.The main contributions of this work are as follows:(1)We propose a simplified transformer architecture that reduces model complexity to a linear relationship with the number of nodes.(2)By integrating the simplified transformer with GNNs,neighborhood information is aggregated while maintaining global attention.The experimental results demonstrate that the proposed model delivers exceptional performance in terms of both accuracy and efficiency.展开更多
As a critical material in construction engineering,concrete requires accurate prediction of its outlet temperature to ensure structural quality and enhance construction efficiency.This study proposes a novel hybrid pr...As a critical material in construction engineering,concrete requires accurate prediction of its outlet temperature to ensure structural quality and enhance construction efficiency.This study proposes a novel hybrid prediction method that integrates a heat conduction physical model with a multilayer perceptron(MLP)neural network,dynamically fused via a weighted strategy to achieve high-precision temperature estimation.Experimental results on an independent test set demonstrated the superior performance of the fused model,with a root mean square error(RMSE)of 1.59℃ and a mean absolute error(MAE)of 1.23℃,representing a 25.3%RMSE reduction compared to conventional physical models.Ambient temperature and coarse aggregate temperature were identified as the most influential variables.Furthermore,the model-based temperature control strategy reduced costs by 0.81 CNY/m^(3),showing significant potential for improving resource efficiency and supporting sustainable construction practices.展开更多
Objective To develop a depression recognition model by integrating the spirit-expression diagnostic framework of traditional Chinese medicine(TCM)with machine learning algorithms.The proposed model seeks to establish ...Objective To develop a depression recognition model by integrating the spirit-expression diagnostic framework of traditional Chinese medicine(TCM)with machine learning algorithms.The proposed model seeks to establish a TCM-informed tool for early depression screening,thereby bridging traditional diagnostic principles with modern computational approaches.Methods The study included patients with depression who visited the Shanghai Pudong New Area Mental Health Center from October 1,2022 to October 1,2023,as well as students and teachers from Shanghai University of Traditional Chinese Medicine during the same period as the healthy control group.Videos of 3–10 s were captured using a Xiaomi Pad 5,and the TCM spirit and expressions were determined by TCM experts(at least 3 out of 5 experts agreed to determine the category of TCM spirit and expressions).Basic information,facial images,and interview information were collected through a portable TCM intelligent analysis and diagnosis device,and facial diagnosis features were extracted using the Open CV computer vision library technology.Statistical analysis methods such as parametric and non-parametric tests were used to analyze the baseline data,TCM spirit and expression features,and facial diagnosis feature parameters of the two groups,to compare the differences in TCM spirit and expression and facial features.Five machine learning algorithms,including extreme gradient boosting(XGBoost),decision tree(DT),Bernoulli naive Bayes(BernoulliNB),support vector machine(SVM),and k-nearest neighbor(KNN)classification,were used to construct a depression recognition model based on the fusion of TCM spirit and expression features.The performance of the model was evaluated using metrics such as accuracy,precision,and the area under the receiver operating characteristic(ROC)curve(AUC).The model results were explained using the Shapley Additive exPlanations(SHAP).Results A total of 93 depression patients and 87 healthy individuals were ultimately included in this study.There was no statistically significant difference in the baseline characteristics between the two groups(P>0.05).The differences in the characteristics of the spirit and expressions in TCM and facial features between the two groups were shown as follows.(i)Quantispirit facial analysis revealed that depression patients exhibited significantly reduced facial spirit and luminance compared with healthy controls(P<0.05),with characteristic features such as sad expressions,facial erythema,and changes in the lip color ranging from erythematous to cyanotic.(ii)Depressed patients exhibited significantly lower values in facial complexion L,lip L,and a values,and gloss index,but higher values in facial complexion a and b,lip b,low gloss index,and matte index(all P<0.05).(iii)The results of multiple models show that the XGBoost-based depression recognition model,integrating the TCM“spirit-expression”diagnostic framework,achieved an accuracy of 98.61%and significantly outperformed four benchmark algorithms—DT,BernoulliNB,SVM,and KNN(P<0.01).(iv)The SHAP visualization results show that in the recognition model constructed by the XGBoost algorithm,the complexion b value,categories of facial spirit,high gloss index,low gloss index,categories of facial expression and texture features have significant contribution to the model.Conclusion This study demonstrates that integrating TCM spirit-expression diagnostic features with machine learning enables the construction of a high-precision depression detection model,offering a novel paradigm for objective depression diagnosis.展开更多
Tomato is a major economic crop worldwide,and diseases on tomato leaves can significantly reduce both yield and quality.Traditional manual inspection is inefficient and highly subjective,making it difficult to meet th...Tomato is a major economic crop worldwide,and diseases on tomato leaves can significantly reduce both yield and quality.Traditional manual inspection is inefficient and highly subjective,making it difficult to meet the requirements of early disease identification in complex natural environments.To address this issue,this study proposes an improved YOLO11-based model,YOLO-SPDNet(Scale Sequence Fusion,Position-Channel Attention,and Dual Enhancement Network).The model integrates the SEAM(Self-Ensembling Attention Mechanism)semantic enhancement module,the MLCA(Mixed Local Channel Attention)lightweight attention mechanism,and the SPA(Scale-Position-Detail Awareness)module composed of SSFF(Scale Sequence Feature Fusion),TFE(Triple Feature Encoding),and CPAM(Channel and Position Attention Mechanism).These enhancements strengthen fine-grained lesion detection while maintaining model lightweightness.Experimental results show that YOLO-SPDNet achieves an accuracy of 91.8%,a recall of 86.5%,and an mAP@0.5 of 90.6%on the test set,with a computational complexity of 12.5 GFLOPs.Furthermore,the model reaches a real-time inference speed of 987 FPS,making it suitable for deployment on mobile agricultural terminals and online monitoring systems.Comparative analysis and ablation studies further validate the reliability and practical applicability of the proposed model in complex natural scenes.展开更多
Rainfall input errors are a major source of uncertainty in flood forecasting,and merging multi-source precipitation data is essential for improving accuracy.Traditional merging methods often prioritize precipitation m...Rainfall input errors are a major source of uncertainty in flood forecasting,and merging multi-source precipitation data is essential for improving accuracy.Traditional merging methods often prioritize precipitation magnitude enhancements while overlooking event detection and false alarms.To address these limitations,this study developed a precipitation integration framework that combines machine learning classification-plus-regression models with Bayesian model averaging(BMA).Three machine learning algorithms-categorical boosting(CatBoost),light gradient boosting machine(LightGBM),and random forest(RF)-were used to improve precipitation event detection.The framework includes spatial unification of raw satellite products using bilinear interpolation,bias correction through classification-plus-regression models,and final merging via a seasonal-scale BMA model.The method integrated GSMaP,IMERG,and PERSIANN satellite precipitation products,with ground observations used for model training(2001-2014)and independent validation(2015-2020)in the Upper Ganjiang River Basin,China.Results showed that the framework significantly enhanced precipitation estimation accuracy and detection capability.LightGBM-based integration exhibited superior detection performance(FAR=0.08,CSI=0.86),while RF-based integration achieved the highest overall accuracy(RMSE=4.67,CC=0.92).Seasonal variations in BMA weights underscored the need to account for seasonal characteristics of precipitation products.Additionally,accuracy improvements were observed across all rainfall categories,especially for heavy rainstorms.The seasonal-scale BMA fusion has combined the strengths of individual corrections and further enhanced precipitation estimation.This research offers a robust method for generating accurate rainfall inputs,providing valuable support for hydrological modeling and flood forecasting applications.展开更多
Reconfigurable array architecture has become an important hardware platform for edge-side deployment of convolutional neural networks due to their high parallelism and flexible programmability.However,traditional mult...Reconfigurable array architecture has become an important hardware platform for edge-side deployment of convolutional neural networks due to their high parallelism and flexible programmability.However,traditional multi-branch convolutional networks suffer from computational redundancy,high memory access overhead,and inefficient branch fusion.Therefore,this paper proposes an adaptive multi-branch convolutional module(AMBC)that integrates software-hardware co-optimization.During training,the learnable fusion coefficients are introduced to enable adaptive fusion of multi-scale features,while in the inference phase,the multiple branches and their normalization parameters are merged with the fusion coefficients into a single 3×3 convolutional kernel through operator fusion.On the SIREA-288 reconfigurable platform,compared with unoptimized multi-branch networks,the proposed AMBC reduces external memory accesses by 47.91%and inference latency by 47.20%,achieving a 1.90×speedup.This approach maximizes the utilization of the reconfigurable logic while minimizing both reconfiguration and data-movement overheads in edge inference.展开更多
Automatic segmentation of landslides from remote sensing imagery is challenging because traditional machine learning and early CNN-based models often fail to generalize across heterogeneous landscapes,where segmentati...Automatic segmentation of landslides from remote sensing imagery is challenging because traditional machine learning and early CNN-based models often fail to generalize across heterogeneous landscapes,where segmentation maps contain sparse and fragmented landslide regions under diverse geographical conditions.To address these issues,we propose a lightweight dual-stream siamese deep learning framework that integrates optical and topographical data fusion with an adaptive decoder,guided multimodal fusion,and deep supervision.The framework is built upon the synergistic combination of cross-attention,gated fusion,and sub-pixel upsampling within a unified dual-stream architecture specifically optimized for landslide segmentation,enabling efficient context modeling and robust feature exchange between modalities.The decoder captures long-range context at deeper levels using lightweight cross-attention and refines spatial details at shallower levels through attention-gated skip fusion,enabling precise boundary delineation and fewer false positives.The gated fusion further enhances multimodal integration of optical and topographical cues,and the deep supervision stabilizes training and improves generalization.Moreover,to mitigate checkerboard artifacts,a learnable sub-pixel upsampling is devised to replace the traditional transposed convolution.Despite its compact design with fewer parameters,the model consistently outperforms state-of-the-art baselines.Experiments on two benchmark datasets,Landslide4Sense and Bijie,confirm the effectiveness of the framework.On the Bijie dataset,it achieves an F1-score of 0.9110 and an intersection over union(IoU)of 0.8839.These results highlight its potential for accurate large-scale landslide inventory mapping and real-time disaster response.The implementation is publicly available at https://github.com/mishaown/DiGATe-UNet-LandSlide-Segmentation(accessed on 3 November 2025).展开更多
To solve the mismatch between the candidate model and the reference model caused by the time change of the tracked head, a novel mean shift algorithm based on a fusion model is provided. A fusion model is employed to ...To solve the mismatch between the candidate model and the reference model caused by the time change of the tracked head, a novel mean shift algorithm based on a fusion model is provided. A fusion model is employed to describe the tracked head by sampling the models of the fore-head and the back-head under different situations. Thus the fusion head reference model is represented by the color distribution estimated from both the fore- head and the back-head. The proposed tracking system is efficient and it is easy to realize the goal of continual tracking of the head by using the fusion model. The results show that the new tracker is robust up to a 360°rotation of the head on a cluttered background and the tracking precision is improved.展开更多
An algorithm of highly maneuvering target tracking is proposed to solve the problem of large tracking error caused by strong maneuver. In this algorithm, a new estimator, named as multi-parameter fusion Singer (MF-Sin...An algorithm of highly maneuvering target tracking is proposed to solve the problem of large tracking error caused by strong maneuver. In this algorithm, a new estimator, named as multi-parameter fusion Singer (MF-Singer) model is derived based on the Singer model and the fuzzy reasoning method by using radial acceleration and velocity of the target, and applied to the problem of maneuvering target tracking in strong maneuvering environment and operating environment. The tracking performance of the MF-Singer model is evaluated and compared with other manuevering tracking models. It is shown that the MF-Singer model outperforms these algorithms in several examples.展开更多
Effect of various spatial and energy distributions of fusion neutron sourceon the calculation of neutron wall loading of Tokamak D-D fusion device has been investigated bymeans of the 3-D Monte Carlo code MCNP. A real...Effect of various spatial and energy distributions of fusion neutron sourceon the calculation of neutron wall loading of Tokamak D-D fusion device has been investigated bymeans of the 3-D Monte Carlo code MCNP. A realistic Monte Carlo source model was developed based onthe accurate representation of the spatial distribution and energy spectrum of fusion neutrons tosolve the complicated problem of tokamak fusion neutron source modelling. The results show thatthose simplified source models will introduce significant uncertainties. For accurate estimation ofthe key nuclear responses of the tokamak design and analyses, the use of the realistic source isrecommended. In addition, the accumulation of tritium produced during D-D plasma operation should becarefully considered.展开更多
In order to meet the demand of testability analysis and evaluation for complex equipment under a small sample test in the equipment life cycle, the hierarchical hybrid testability model- ing and evaluation method （HH...In order to meet the demand of testability analysis and evaluation for complex equipment under a small sample test in the equipment life cycle, the hierarchical hybrid testability model- ing and evaluation method （HHTME）, which combines the testabi- lity structure model （TSM） with the testability Bayesian networks model （TBNM）, is presented. Firstly, the testability network topo- logy of complex equipment is built by using the hierarchical hybrid testability modeling method. Secondly, the prior conditional prob- ability distribution between network nodes is determined through expert experience. Then the Bayesian method is used to update the conditional probability distribution, according to history test information, virtual simulation information and similar product in- formation. Finally, the learned hierarchical hybrid testability model （HHTM） is used to estimate the testability of equipment. Compared with the results of other modeling methods, the relative deviation of the HHTM is only 0.52%, and the evaluation result is the most accu rate.展开更多
This paper presents a data fusion algorithm for dynamic system with multi-sensor and uncertain system models. The algorithm is mainly based on Kalman filter and interacting multiple model(IMM). It processes crosscorre...This paper presents a data fusion algorithm for dynamic system with multi-sensor and uncertain system models. The algorithm is mainly based on Kalman filter and interacting multiple model(IMM). It processes crosscorrelated sensor noises by using augmented fusion before model interacting. And eigenvalue decomposition is utilized to reduce calculation complexity and implement parallel computing. In simulation part, the feasibility of the algorithm was tested and verified, and the relationship between sensor number and the estimation precision was studied. Results show that simply increasing the number of sensor cannot always improve the performance of the estimation. Type and number of sensors should be optimized in practical applications.展开更多
Dezert-Smarandache（DSm） theory, a new information fusion theory, is widely applied in image processing, multiple targets tracking identification, and other areas for its excellent processing ability of imperfect inf...Dezert-Smarandache（DSm） theory, a new information fusion theory, is widely applied in image processing, multiple targets tracking identification, and other areas for its excellent processing ability of imperfect information. However, earlier research on DSm theory mainly focused on one sort of questions. An evidence fusion procedure is proposed based on the hybrid DSm model to compensate for a lack of research on the entire information procedure of DSm theory. This paper analyzes the evidence fusion procedure, as well as correlative node input and output information. Key steps and detailed procedures of evidence fusion are also discussed. Finally, an experiment illustrates the efficiency of the proposed evidence fusion procedure.展开更多
Lithium-ion batteries(LIBs)have emerged as the preferred energy storage systems for various types of electric transports,including electric vehicles,electric boats,electric trains,and electric airplanes.The energy man...Lithium-ion batteries(LIBs)have emerged as the preferred energy storage systems for various types of electric transports,including electric vehicles,electric boats,electric trains,and electric airplanes.The energy management of LIBs in electric transports for all-climate and long-life operation requires the accurate estimation of state of charge(SOC)and capacity in real-time.This study proposes a multistage model fusion algorithm to co-estimate SOC and capacity.Firstly,based on the assumption of a normal distribution,the mean and variance of the residual error from the model at different ageing levels are used to calculate the weight for the establishment of a fusion model with stable parameters.Secondly,a differential error gain with forward-looking ability is introduced into a proportional–integral observer(PIO)to accelerate convergence speed.Thirdly,a fusion algorithm is developed by combining a multistage model and proportional–integral–differential observer(PIDO)to co-estimate SOC and capacity under a complex application environment.Fourthly,the convergence and anti-noise performance of the fusion algorithm are discussed.Finally,the hardware-in-the-loop platform is set up to verify the performance of the fusion algorithm.The validation results of different aged LIBs over a wide range of temperature show that the presented fusion algorithm can realize a high-accuracy estimation of SOC and capacity with the relative errors within 2%and 3.3%,respectively.展开更多
基金supported by the National Natural Science Foundation of China(No.41971339)the SDUST Research Fund(No.2019TDJH103)。
文摘The rapid melting of Arctic sea ice poses significant risks to the safety of shipping routes.Accurate remote sensing data on sea ice concentration(SIC)is crucial for effective route planning of ships and ensuring navigational safety.Despite the availability of numerous SIC products in China,these datasets still lag behind mainstream international products in terms of data accuracy,spatiotemporal resolution,and time span.To enhance the accuracy of China's domestic SIC remote sensing data,this study used the SIC data derived from the passive microwave remote sensing dataset provided by the University of Bremen(BRM-SIC)as a reference to conduct a comprehensive evaluation and analysis of two additional SIC datasets:the dataset derived from the microwave radiation imager(MWRI)aboard the FY-3D satellite,provided by the National Satellite Meteorological Center(FY-SIC),and the dataset obtained through the DT-ASI algorithm from the microwave imager of the FY-3D satellite,provided by Ocean University of China(OUC-SIC).Based on the evaluation results,a TransUnet fusion correction model was developed.The performance of this model was then compared against Ordinary Least Squares(OLS),Random Forest(RF),and UNet correction models,through spatial and temporal analyses.Results indicate that,compared to FY-SIC data,the RMSE of the OUC-SIC data and the standard data is reduced by24.245%,while the R is increased by 12.516%.Overall,the accuracy of OUC-SIC data is superior to that of FY-SIC data.During the research period(2020–2022),the standard deviation(SD)and coefficient of variation(CV)of OUC-SIC were 3.877%and 10.582%,respectively,while those for FY-SIC were 7.836%and 7.982%,respectively.In the study area,compared with OUC-SIC data,FYSIC data exhibited a larger standard deviation of deviation and a smaller coefficient of variation of deviation across most sea areas.These results indicate that the OUC-SIC data exhibit better temporal and spatial stability,whereas the FY-SIC data show stronger relative dimensionless stability.Among the four correction models,all showed improvements over the original,unfused corrected data.The fusion corrections using the OLS,RF,UNet,and TransUnet models reduced RMSE by 5.563%,14.601%,42.927%,and48.316%,respectively.Correspondingly,R increased by 0.463%,1.176%,3.951%,and 4.342%,respectively.Among these models,TransUnet performed the best,effectively integrating the advantages of FY-SIC and OUC-SIC data and notably improving the overall accuracy and spatiotemporal stability of SIC data.
基金funded by National Natural Science Foundation of China Key Program(12431014)Key Project of Hunan Education Department(22A0126)+1 种基金Natural Science Foundation of Hunan Province(2022JJ30555)Postgraduate Scientific Research Innovation Project of Xiangtan University(XDCX2024Y172)。
文摘Tropospheric zenith wet delay(ZWD)plays a vital role in the analysis of space geodetic observations.In recent years,machine learning methods have been increasingly applied to improve the accuracy of ZWD calculations.However,a single machine learning model has limited generalization capabilities.To address these limitations,this study introduces a novel machine learning fusion(MLF)algorithm with stronger generalization capabilities to enhance ZWD modeling and prediction accuracy.The MLF algorithm utilizes a two-layer structure integrating extra trees(ET),backpropagation neural network(BPNN),and linear regression models.By comparing the root mean square error(RMSE)of these models,we found that both ET-based and MLF-based models outperform RF-based and BPNN-based models in terms of internal and external accuracy,across both surface meteorological data-based and blind models.The improvement in exte rnal accuracy is particularly significant in the blind models.Our re sults show that the MLF(with an RMSE of 3.93 cm)and ET(3.99 cm)models outperform the traditional GPT3model(4.07 cm),while the RF(4.21 cm)and BPNN(4.14 cm)have worse external accuracies than the GPT3 model.It is worth noting that the BPNN suffered from overfitting during external accuracy tests,which was avoided by the MLF.In summary,regardless of the availability of surface meteorological data,the MLF-based empirical models demonstrate superior internal and external accuracy compared to the other tested models in this study.
基金supported by the National Key R&D Program of China (Grant No.2022YFF0503700)the National Natural Science Foundation of China (42074196, 41925018)
文摘Solar flare prediction is an important subject in the field of space weather.Deep learning technology has greatly promoted the development of this subject.In this study,we propose a novel solar flare forecasting model integrating Deep Residual Network(ResNet)and Support Vector Machine(SVM)for both≥C-class(C,M,and X classes)and≥M-class(M and X classes)flares.We collected samples of magnetograms from May 1,2010 to September 13,2018 from Space-weather Helioseismic and Magnetic Imager(HMI)Active Region Patches and then used a cross-validation method to obtain seven independent data sets.We then utilized five metrics to evaluate our fusion model,based on intermediate-output extracted by ResNet and SVM using the Gaussian kernel function.Our results show that the primary metric true skill statistics(TSS)achieves a value of 0.708±0.027 for≥C-class prediction,and of 0.758±0.042 for≥M-class prediction;these values indicate that our approach performs significantly better than those of previous studies.The metrics of our fusion model’s performance on the seven datasets indicate that the model is quite stable and robust,suggesting that fusion models that integrate an excellent baseline network with SVM can achieve improved performance in solar flare prediction.Besides,we also discuss the performance impact of architectural innovation in our fusion model.
基金co-supported by Basic and Applied Basic Research Foundation of Guangdong Province(No.2022A1515110622)Natural Science Basic Research Program of Shaanxi Province(No.2023-JC-QN-0548)+1 种基金National Key R&D Program of China(No.2022YFB3402200)the Fundamental Research Funds for the Central Universities。
文摘This study presents new methods to effectively model the anisotropic yielding and hardening behavior of laser powder bed fusion fabricated aluminum alloy under both monotonic and cyclic loading conditions.The proposed model combines the yield surface-interpolation method to accurately describe the anisotropic hardening rates in various directions,with the Chaboche kinematic hardening rule to precisely reflect the cyclic characteristics.For numerical implementation of the combined anisotropic and cyclic constitutive model,a fully implicit stress integration algorithm based on return mapping method is provided.Moreover,the multiple parameters associated with the model are categorized and identified in an uncoupled manner.The isotropic and cyclic hardening parameters are determined by an inverse method,and the stability of the optimization outcomes is validated by applying different starting points for the parameters.Particularly,the back-stress effect on the identification of anisotropic parameters associated with the stress invariant-based Hill48 yield function is considered for the first time.This consideration leads to an improved prediction accuracy compared to the identification of anisotropic parameters without considering back-stress effect.The combined anisotropic and cyclic constitutive model,along with the calibrated parameters,are proven capable of accurately reproducing the intricate deformation behavior of laser powder bed fusion fabricated AlSi10Mg.
文摘Semi-crystalline polymer laser powder bed fusion(L-PBF)has recently attracted increasing interest due to its potential for fabricating complex geometry.However,a more comprehensive understanding of the underlying physics during L-PBF is required to better control the properties of the final part.This work proposed a multi-layer numerical model to study the temperature and phase evolution during the polyamide-12(PA12)L-PBF process.The Descend and Parallel Chord methods were introduced to improve the convergence of the non-linear thermal solver.The level-set-based mesh adaptation strategy,governed by multi-physical fields,was applied to alleviate the calculation and accurately track the phase evolution.The processing simulation on the dog-bone model revealed that preheating temperature significantly influences the crystallization behavior.Finally,the multi-layer simulation demonstrated that such a developed numerical model can be used to study the phase transformation during powder layer updating and the cyclic laser sintering phenomena.Moreover,the numerical study suggested that crystallization occurs slowly during the L-PBF process.
文摘With the advancement of scientific research and the rapid growth of the internet,academic users increasingly face challenges in obtaining accurate information about peer research.As a key component of big data analytics,user profiling has emerged as a critical focus in the scientific research community.While graph neural networks(GNNs)perform well in various graph learning tasks,their scalability to large graphs becomes problematic as the number of nodes increases due to computational complexity.To address this issue,this study proposes a novel academic user profiling model based on graph neural networks tailored to the unique characteristics of scientific research networks.The main contributions of this work are as follows:(1)We propose a simplified transformer architecture that reduces model complexity to a linear relationship with the number of nodes.(2)By integrating the simplified transformer with GNNs,neighborhood information is aggregated while maintaining global attention.The experimental results demonstrate that the proposed model delivers exceptional performance in terms of both accuracy and efficiency.
基金funded by National Key Research and Development Plan(2018YFC0406703)Supported by the National Natural Science Foundation of China(51779277)+4 种基金Chinese Academy of Water Sciences(SD0145B072021)Supported by the State Key Laboratory of Flow Water Cycle Simulation and Regulation,SKL2022ZD05Support provided by the fund of State Key Laboratory of Water Cycle and Water Security,IWHR(Grant No.SKL2024YJZD05)Support provided by the fund of Power China(DJ-ZDXM-2020-50)Support provided by the fund of Research and Application of Intelligent Simulation and Intelligent Control Technology for Structural States of Gravity DAMS in Jingling Reservoir Project,Zhejiang Province(JLSKFW-2024113).
文摘As a critical material in construction engineering,concrete requires accurate prediction of its outlet temperature to ensure structural quality and enhance construction efficiency.This study proposes a novel hybrid prediction method that integrates a heat conduction physical model with a multilayer perceptron(MLP)neural network,dynamically fused via a weighted strategy to achieve high-precision temperature estimation.Experimental results on an independent test set demonstrated the superior performance of the fused model,with a root mean square error(RMSE)of 1.59℃ and a mean absolute error(MAE)of 1.23℃,representing a 25.3%RMSE reduction compared to conventional physical models.Ambient temperature and coarse aggregate temperature were identified as the most influential variables.Furthermore,the model-based temperature control strategy reduced costs by 0.81 CNY/m^(3),showing significant potential for improving resource efficiency and supporting sustainable construction practices.
基金General Program of National Natural Science Foundation of China(82474390)Construction Project of Pudong New Area Famous TCM Studios(National Pilot Zone for TCM Development,Shanghai)(PDZY-2025-0716)Shanghai Municipal Science and Technology Program Project Shanghai Key Laboratory of Health Identification and Assessment(21DZ2271000).
文摘Objective To develop a depression recognition model by integrating the spirit-expression diagnostic framework of traditional Chinese medicine(TCM)with machine learning algorithms.The proposed model seeks to establish a TCM-informed tool for early depression screening,thereby bridging traditional diagnostic principles with modern computational approaches.Methods The study included patients with depression who visited the Shanghai Pudong New Area Mental Health Center from October 1,2022 to October 1,2023,as well as students and teachers from Shanghai University of Traditional Chinese Medicine during the same period as the healthy control group.Videos of 3–10 s were captured using a Xiaomi Pad 5,and the TCM spirit and expressions were determined by TCM experts(at least 3 out of 5 experts agreed to determine the category of TCM spirit and expressions).Basic information,facial images,and interview information were collected through a portable TCM intelligent analysis and diagnosis device,and facial diagnosis features were extracted using the Open CV computer vision library technology.Statistical analysis methods such as parametric and non-parametric tests were used to analyze the baseline data,TCM spirit and expression features,and facial diagnosis feature parameters of the two groups,to compare the differences in TCM spirit and expression and facial features.Five machine learning algorithms,including extreme gradient boosting(XGBoost),decision tree(DT),Bernoulli naive Bayes(BernoulliNB),support vector machine(SVM),and k-nearest neighbor(KNN)classification,were used to construct a depression recognition model based on the fusion of TCM spirit and expression features.The performance of the model was evaluated using metrics such as accuracy,precision,and the area under the receiver operating characteristic(ROC)curve(AUC).The model results were explained using the Shapley Additive exPlanations(SHAP).Results A total of 93 depression patients and 87 healthy individuals were ultimately included in this study.There was no statistically significant difference in the baseline characteristics between the two groups(P>0.05).The differences in the characteristics of the spirit and expressions in TCM and facial features between the two groups were shown as follows.(i)Quantispirit facial analysis revealed that depression patients exhibited significantly reduced facial spirit and luminance compared with healthy controls(P<0.05),with characteristic features such as sad expressions,facial erythema,and changes in the lip color ranging from erythematous to cyanotic.(ii)Depressed patients exhibited significantly lower values in facial complexion L,lip L,and a values,and gloss index,but higher values in facial complexion a and b,lip b,low gloss index,and matte index(all P<0.05).(iii)The results of multiple models show that the XGBoost-based depression recognition model,integrating the TCM“spirit-expression”diagnostic framework,achieved an accuracy of 98.61%and significantly outperformed four benchmark algorithms—DT,BernoulliNB,SVM,and KNN(P<0.01).(iv)The SHAP visualization results show that in the recognition model constructed by the XGBoost algorithm,the complexion b value,categories of facial spirit,high gloss index,low gloss index,categories of facial expression and texture features have significant contribution to the model.Conclusion This study demonstrates that integrating TCM spirit-expression diagnostic features with machine learning enables the construction of a high-precision depression detection model,offering a novel paradigm for objective depression diagnosis.
基金Tianmin Tianyuan Boutique Vegetable Industry Technology Service Station(Grant No.2024120011003081)Development of Environmental Monitoring and Traceability System for Wuqing Agricultural Production Areas(Grant No.2024120011001866)。
文摘Tomato is a major economic crop worldwide,and diseases on tomato leaves can significantly reduce both yield and quality.Traditional manual inspection is inefficient and highly subjective,making it difficult to meet the requirements of early disease identification in complex natural environments.To address this issue,this study proposes an improved YOLO11-based model,YOLO-SPDNet(Scale Sequence Fusion,Position-Channel Attention,and Dual Enhancement Network).The model integrates the SEAM(Self-Ensembling Attention Mechanism)semantic enhancement module,the MLCA(Mixed Local Channel Attention)lightweight attention mechanism,and the SPA(Scale-Position-Detail Awareness)module composed of SSFF(Scale Sequence Feature Fusion),TFE(Triple Feature Encoding),and CPAM(Channel and Position Attention Mechanism).These enhancements strengthen fine-grained lesion detection while maintaining model lightweightness.Experimental results show that YOLO-SPDNet achieves an accuracy of 91.8%,a recall of 86.5%,and an mAP@0.5 of 90.6%on the test set,with a computational complexity of 12.5 GFLOPs.Furthermore,the model reaches a real-time inference speed of 987 FPS,making it suitable for deployment on mobile agricultural terminals and online monitoring systems.Comparative analysis and ablation studies further validate the reliability and practical applicability of the proposed model in complex natural scenes.
基金supported by the National Natural Science Foundation of China(42471049).
文摘Rainfall input errors are a major source of uncertainty in flood forecasting,and merging multi-source precipitation data is essential for improving accuracy.Traditional merging methods often prioritize precipitation magnitude enhancements while overlooking event detection and false alarms.To address these limitations,this study developed a precipitation integration framework that combines machine learning classification-plus-regression models with Bayesian model averaging(BMA).Three machine learning algorithms-categorical boosting(CatBoost),light gradient boosting machine(LightGBM),and random forest(RF)-were used to improve precipitation event detection.The framework includes spatial unification of raw satellite products using bilinear interpolation,bias correction through classification-plus-regression models,and final merging via a seasonal-scale BMA model.The method integrated GSMaP,IMERG,and PERSIANN satellite precipitation products,with ground observations used for model training(2001-2014)and independent validation(2015-2020)in the Upper Ganjiang River Basin,China.Results showed that the framework significantly enhanced precipitation estimation accuracy and detection capability.LightGBM-based integration exhibited superior detection performance(FAR=0.08,CSI=0.86),while RF-based integration achieved the highest overall accuracy(RMSE=4.67,CC=0.92).Seasonal variations in BMA weights underscored the need to account for seasonal characteristics of precipitation products.Additionally,accuracy improvements were observed across all rainfall categories,especially for heavy rainstorms.The seasonal-scale BMA fusion has combined the strengths of individual corrections and further enhanced precipitation estimation.This research offers a robust method for generating accurate rainfall inputs,providing valuable support for hydrological modeling and flood forecasting applications.
基金Supported by the National Science and Technology Major Project of China(2022ZD0119005)the Natural Science Project of Shaanxi Province(2025JC-YBMS-754,2024JC-YBMS-539)。
文摘Reconfigurable array architecture has become an important hardware platform for edge-side deployment of convolutional neural networks due to their high parallelism and flexible programmability.However,traditional multi-branch convolutional networks suffer from computational redundancy,high memory access overhead,and inefficient branch fusion.Therefore,this paper proposes an adaptive multi-branch convolutional module(AMBC)that integrates software-hardware co-optimization.During training,the learnable fusion coefficients are introduced to enable adaptive fusion of multi-scale features,while in the inference phase,the multiple branches and their normalization parameters are merged with the fusion coefficients into a single 3×3 convolutional kernel through operator fusion.On the SIREA-288 reconfigurable platform,compared with unoptimized multi-branch networks,the proposed AMBC reduces external memory accesses by 47.91%and inference latency by 47.20%,achieving a 1.90×speedup.This approach maximizes the utilization of the reconfigurable logic while minimizing both reconfiguration and data-movement overheads in edge inference.
基金funded by the National Natural Science Foundation of China,grant number 62262045the Fundamental Research Funds for the Central Universities,grant number 2023CDJYGRH-YB11the Open Funding of SUGON Industrial Control and Security Center,grant number CUIT-SICSC-2025-03.
文摘Automatic segmentation of landslides from remote sensing imagery is challenging because traditional machine learning and early CNN-based models often fail to generalize across heterogeneous landscapes,where segmentation maps contain sparse and fragmented landslide regions under diverse geographical conditions.To address these issues,we propose a lightweight dual-stream siamese deep learning framework that integrates optical and topographical data fusion with an adaptive decoder,guided multimodal fusion,and deep supervision.The framework is built upon the synergistic combination of cross-attention,gated fusion,and sub-pixel upsampling within a unified dual-stream architecture specifically optimized for landslide segmentation,enabling efficient context modeling and robust feature exchange between modalities.The decoder captures long-range context at deeper levels using lightweight cross-attention and refines spatial details at shallower levels through attention-gated skip fusion,enabling precise boundary delineation and fewer false positives.The gated fusion further enhances multimodal integration of optical and topographical cues,and the deep supervision stabilizes training and improves generalization.Moreover,to mitigate checkerboard artifacts,a learnable sub-pixel upsampling is devised to replace the traditional transposed convolution.Despite its compact design with fewer parameters,the model consistently outperforms state-of-the-art baselines.Experiments on two benchmark datasets,Landslide4Sense and Bijie,confirm the effectiveness of the framework.On the Bijie dataset,it achieves an F1-score of 0.9110 and an intersection over union(IoU)of 0.8839.These results highlight its potential for accurate large-scale landslide inventory mapping and real-time disaster response.The implementation is publicly available at https://github.com/mishaown/DiGATe-UNet-LandSlide-Segmentation(accessed on 3 November 2025).
基金The National Natural Science Foundation of China(No.60672094,60673188,U0735004)the National High Technology Research and Development Program of China(863 Program)(No.2008AA01Z303)the National Basic Research Program of China (973 Program)(No.2009CB320804)
文摘To solve the mismatch between the candidate model and the reference model caused by the time change of the tracked head, a novel mean shift algorithm based on a fusion model is provided. A fusion model is employed to describe the tracked head by sampling the models of the fore-head and the back-head under different situations. Thus the fusion head reference model is represented by the color distribution estimated from both the fore- head and the back-head. The proposed tracking system is efficient and it is easy to realize the goal of continual tracking of the head by using the fusion model. The results show that the new tracker is robust up to a 360°rotation of the head on a cluttered background and the tracking precision is improved.
基金supported by the National Natural Science Foundation of China(6153102061471383)
文摘An algorithm of highly maneuvering target tracking is proposed to solve the problem of large tracking error caused by strong maneuver. In this algorithm, a new estimator, named as multi-parameter fusion Singer (MF-Singer) model is derived based on the Singer model and the fuzzy reasoning method by using radial acceleration and velocity of the target, and applied to the problem of maneuvering target tracking in strong maneuvering environment and operating environment. The tracking performance of the MF-Singer model is evaluated and compared with other manuevering tracking models. It is shown that the MF-Singer model outperforms these algorithms in several examples.
基金The project supported partly by the National Science Foundation of Anhui Province (No. 0104360)
文摘Effect of various spatial and energy distributions of fusion neutron sourceon the calculation of neutron wall loading of Tokamak D-D fusion device has been investigated bymeans of the 3-D Monte Carlo code MCNP. A realistic Monte Carlo source model was developed based onthe accurate representation of the spatial distribution and energy spectrum of fusion neutrons tosolve the complicated problem of tokamak fusion neutron source modelling. The results show thatthose simplified source models will introduce significant uncertainties. For accurate estimation ofthe key nuclear responses of the tokamak design and analyses, the use of the realistic source isrecommended. In addition, the accumulation of tritium produced during D-D plasma operation should becarefully considered.
基金supported by the National Defense Pre-research Foundation of China(51327030104)
文摘In order to meet the demand of testability analysis and evaluation for complex equipment under a small sample test in the equipment life cycle, the hierarchical hybrid testability model- ing and evaluation method （HHTME）, which combines the testabi- lity structure model （TSM） with the testability Bayesian networks model （TBNM）, is presented. Firstly, the testability network topo- logy of complex equipment is built by using the hierarchical hybrid testability modeling method. Secondly, the prior conditional prob- ability distribution between network nodes is determined through expert experience. Then the Bayesian method is used to update the conditional probability distribution, according to history test information, virtual simulation information and similar product in- formation. Finally, the learned hierarchical hybrid testability model （HHTM） is used to estimate the testability of equipment. Compared with the results of other modeling methods, the relative deviation of the HHTM is only 0.52%, and the evaluation result is the most accu rate.
基金the National Natural Science Foundation of China(No.61374160)the Shanghai Aerospace Science and Technology Innovation Fund(No.SAST201237)
文摘This paper presents a data fusion algorithm for dynamic system with multi-sensor and uncertain system models. The algorithm is mainly based on Kalman filter and interacting multiple model(IMM). It processes crosscorrelated sensor noises by using augmented fusion before model interacting. And eigenvalue decomposition is utilized to reduce calculation complexity and implement parallel computing. In simulation part, the feasibility of the algorithm was tested and verified, and the relationship between sensor number and the estimation precision was studied. Results show that simply increasing the number of sensor cannot always improve the performance of the estimation. Type and number of sensors should be optimized in practical applications.
基金supported by the National Natural Science Foundation of China(61102168)the Military Innovation Foundation(X11QN106)
文摘Dezert-Smarandache（DSm） theory, a new information fusion theory, is widely applied in image processing, multiple targets tracking identification, and other areas for its excellent processing ability of imperfect information. However, earlier research on DSm theory mainly focused on one sort of questions. An evidence fusion procedure is proposed based on the hybrid DSm model to compensate for a lack of research on the entire information procedure of DSm theory. This paper analyzes the evidence fusion procedure, as well as correlative node input and output information. Key steps and detailed procedures of evidence fusion are also discussed. Finally, an experiment illustrates the efficiency of the proposed evidence fusion procedure.
基金This work was supported by the National Key Research and Development Program of China(2017YFB0103802)the National Natural Science Foundation of China(51922006 and 51707011).
文摘Lithium-ion batteries(LIBs)have emerged as the preferred energy storage systems for various types of electric transports,including electric vehicles,electric boats,electric trains,and electric airplanes.The energy management of LIBs in electric transports for all-climate and long-life operation requires the accurate estimation of state of charge(SOC)and capacity in real-time.This study proposes a multistage model fusion algorithm to co-estimate SOC and capacity.Firstly,based on the assumption of a normal distribution,the mean and variance of the residual error from the model at different ageing levels are used to calculate the weight for the establishment of a fusion model with stable parameters.Secondly,a differential error gain with forward-looking ability is introduced into a proportional–integral observer(PIO)to accelerate convergence speed.Thirdly,a fusion algorithm is developed by combining a multistage model and proportional–integral–differential observer(PIDO)to co-estimate SOC and capacity under a complex application environment.Fourthly,the convergence and anti-noise performance of the fusion algorithm are discussed.Finally,the hardware-in-the-loop platform is set up to verify the performance of the fusion algorithm.The validation results of different aged LIBs over a wide range of temperature show that the presented fusion algorithm can realize a high-accuracy estimation of SOC and capacity with the relative errors within 2%and 3.3%,respectively.
