In the era of materials genome engineering,data-driven machine learning has become a powerful tool for accelerating the re-search and development of metallic materials.However,the predictive accuracy and generalizatio...In the era of materials genome engineering,data-driven machine learning has become a powerful tool for accelerating the re-search and development of metallic materials.However,the predictive accuracy and generalization ability of traditional machine learning models are often limited by the scarcity and heterogeneity of available data,especially in small-sample scenarios.To address these chal-lenges,transfer learning has emerged as an effective strategy to leverage knowledge from related domains,thereby enhancing model per-formance with limited target data.This review systematically summarizes the fundamental concepts,methodologies,and representative applications of transfer learning in the prediction of metallic materials'properties.Transfer learning can be categorized into feature-based,instance-based,parameter-based,and knowledge-based methods.This work discusses their respective mechanisms,advantages,and limit-ations.Case studies demonstrate that transfer learning can significantly improve prediction accuracy,data efficiency,and model inter-pretability in tasks such as mechanical property prediction and alloy design.Furthermore,this work highlights emerging trends including hybrid,multi-task,meta,and adaptive transfer learning,which further expand the applicability of these techniques.Finally,this work out-lines future research directions,emphasizing the need for data standardization,algorithmic innovation,multimodal data fusion,and the in-tegration of physical principles to achieve robust,interpretable,and generalizable models.The perspectives presented aim to advance the intelligent design and discovery of metallic materials,promoting efficient knowledge transfer and collaborative innovation in materials science.展开更多
In practical microgrids,current saturation of inverters and power interaction coupling of different forms of DERs complicate the system's transient behaviors.Existing methods of online transient stability predicti...In practical microgrids,current saturation of inverters and power interaction coupling of different forms of DERs complicate the system's transient behaviors.Existing methods of online transient stability prediction(TSP)are suitable for power systems consisting of homogeneous distributed energy resources(DERs),thus showing limited accuracy for stability prediction of microgrids.This paper develops a deep-learning-based TSP method for accurate online prediction of microgrids consisting of diverse forms of DERs under current saturation.First,a general key input feature selection method for microgrid TSP is systematically designed to ensure prediction accuracy.It is derived from a comprehensive mechanism analysis of the influence of DER's intrinsic and interaction characteristics under current saturation.Besides,impacts of load fluctuation and fault change are also considered to improve robust prediction performance.Second,to further improve prediction accuracy,an online TSP model based on deep learning is developed by effectively using the powerful nonlinear mapping capability of the deep belief network(DBN).Then,by combining feature selection method and deep-learning-based TSP model,an online TSP method is derived.Test results show the proposed method greatly improves accuracy of microgrid TSP under complex operating conditions.Furthermore,the method effectively avoids feature redundancy and the curse of dimensionality.Numbers of input features are independent of the scale of microgrids.展开更多
Deep learning and fatigue life prediction remain focal research areas in rail vehicle engineering.This study addresses the vibration fatigue of wheelset lifting lug in Chengdu Metro Line 1 bogies,aiming to develop a f...Deep learning and fatigue life prediction remain focal research areas in rail vehicle engineering.This study addresses the vibration fatigue of wheelset lifting lug in Chengdu Metro Line 1 bogies,aiming to develop a fatigue life prediction method for critical bogie components using deep learning models and measured track load spectra.Extensive field tests on Chengdu Metro Line 1 were conducted to acquire acceleration and stress response data of the wheelset lifting lug,generating training samples for the neural network system.Component stress responses were calculated via time-domain track acceleration and validated against in-situ stress measurements.Results show that neural network-fitted dynamic stress values exhibit excellent consistency with measured data,with errors constrained within 5%.This study validates the proposed small-sample deep learning approach as an effective and accurate solution for fatigue life prediction of critical bogie components under operational load conditions.展开更多
This study presents a hybrid methodology for predicting building collapses within the Intelligent Circular Resilience(ICR)framework.This uses a supervised Machine Learning(ML)approach,earthquake damage re-ports,and th...This study presents a hybrid methodology for predicting building collapses within the Intelligent Circular Resilience(ICR)framework.This uses a supervised Machine Learning(ML)approach,earthquake damage re-ports,and the Simplified Resilience Index(SRI),derived from existing earthquake damage models(EDM)-based on fragility and vulnerability functions-used in the probabilistic seismic risk assessment(PSRA).A curated building damage database comprising 89 structures(71 collapsed and 18 non-collapsed)from ten countries affected by major earthquakes(Mw 6.1-8.1,epicentral distances of 3-125 km,and PGA values ranging from 0.14 g to 0.82 g)was developed,including attributes related to exposure:occupancy,main structural material,number of stories,construction year,and hazard:magnitude,epicentral distance,intensity measures(Peak-ground acceleration,PGA,and elastic spectral acceleration).The dataset includes events such as the 2017 Puebla-Morelos earthquake(Mw 7.1,Mexico),the 1999 Kocaeli earthquake(Mw 7.6,Turkey),and the 2011 Christchurch earthquake(Mw 6.1,New Zealand),among others.Likewise,dependent attributes such as time elapsed and SRI(under 120-,180-,and 365-day recovery scenarios)were calculated using 2-EDMs.Eight Random Forest models were trained and tested for collapse and non-collapse classification using combinations of independent and dependent attributes.The results indicate that models incorporating exposure-related varia-bles-such as structural material,number of stories,construction year,and occupancy-alongside the SRI significantly improve collapse classification performance,achieving recall and F1 scores above 95%.Notably,many collapsed buildings exhibited low intensities(PGA≤0.25 g),emphasizing the influence of local site effects-particularly in Mexico City.The findings demonstrate that incorporating SRI enhances the reliability of collapse prediction and supports its use as an interpretable resilience proxy during early ICR stages.This hybrid methodology bridges empirical data,traditional PSRA models,and ML techniques,contributing to more accurate and scalable post-earthquake resilience assessments.展开更多
Hard disk drives(HDDs)serve as the primary storage devices in modern data centers.Once a failure occurs,it often leads to severe data loss,significantly degrading the reliability of storage systems.Numerous studies ha...Hard disk drives(HDDs)serve as the primary storage devices in modern data centers.Once a failure occurs,it often leads to severe data loss,significantly degrading the reliability of storage systems.Numerous studies have proposed machine learning-based HDD failure prediction models.However,the Self-Monitoring,Analysis,and Reporting Technology(SMART)attributes differ across HDD manufacturers.We define hard drives of the same brand and model as homogeneous HDD groups,and those from different brands or models as heterogeneous HDD groups.In practical engineering scenarios,a data center is often composed of a heterogeneous population of HDDs,spanning multiple vendors and models.Existing research predominantly focuses on homogeneous datasets,ignoring the model’s generalization capability across heterogeneous HDDs.As a result,HDD models with limited samples often suffer from poor training effectiveness and prediction performance.To address this issue,we investigate generalizable SMART predictors across heterogeneous HDD groups.By extracting time-series features within a fixed sliding time window,we propose a Heterogeneous Disk Failure Prediction Method based on Time Series Features(HDFPM)framework.This method is adaptable to HDD models with limited sample sizes,thereby enhancing its applicability and robustness across diverse drive populations.Experimental results show that the proposed model achieves an F1-score of 0.9518 when applied to two different Seagate HDD models,while maintaining the False Positive Rate(FPR)below 1%.After incorporating the Complexity-Ratio Dynamic Time Warping(CDTW)based feature enhancement method,the best prediction model achieves a True Positive Rate(TPR)of up to 0.93 between the two models.For next-day failure prediction across various Seagate models,the model achieves an F1-score of up to 0.8792.Moreover,the experimental results also show that within the same brand,the higher the proportion of shared SMART attributes across different models,the better the prediction performance.In addition,HDFPMdemonstrates the best stability andmost significant performance in heterogeneous environments.展开更多
To address the challenges of complex fluvial sandbody distribution and difficult remaining oil recovery in mature continental oilfields,this study focuses on key issues in reservoir identification such as ambiguous na...To address the challenges of complex fluvial sandbody distribution and difficult remaining oil recovery in mature continental oilfields,this study focuses on key issues in reservoir identification such as ambiguous narrow-channel boundaries and subdivision of multi-stage superimposed sandbodies.Taking the Upper Cretaceous continental sandstone in the Sazhong Oilfield of the Daqing Placanticline as an example,a technical system integrating OVT high-resolution processing,multi-attribute fusion,and varible-scale inversion was developed to establish a complete workflow from seismic processing to reservoir prediction and remaining oil recovery.The following results are obtained.First,the Offset Vector Tile(OVT)seismic processing technology is extended,for the first time,from fracture imaging to sandbody prediction,in order to address the weak seismic responses from boundaries of narrow and thin sandbodies.A geology-oriented OVT partitioning method is developed to significantly improve the imaging accuracy,enabling identification of channel sandbodies as narrow as 50 m.Second,an amplitude-coherence dual-attribute fusion method is proposed for predicting narrow channel boundaries between wells.Constrained by a sedimentary unit-level sequence chronostratigraphic framework,this method accurately delineates 800-2000 m long subaqueous distributary channels with bifurcation-convergence features.Third,considering the superimposition of multi-stage channels,a three-level variable-scale stratigraphic model(sandstone groups,sublayers,sedimentary units)is constructed to overcome single-scale modeling limitations,successfully characterizing key sedimentary features like meandering river“cut-offs”through 3D seismic inversion.Based on these advances,a direct link between seismic prediction and remaining oil recovery is established.The horizontal wells deployed using narrow-channel predictions encountered oil-bearing sandstones in the horizontal section by 97%,and achieved initial daily production of 12.5 t per well.Precise identification of individual channel boundaries within 17 composite sandbodies guided recovery processes in 135 wells,yielding an average daily increase of 2.8 t per well and a cumulative increase of 13.6×10^(4)t.展开更多
Black-box models have demonstrated remarkable accuracy in forecasting building energy loads.However,they usually lack interpretability and do not incorporate domain knowledge,making it difficult for users to trust the...Black-box models have demonstrated remarkable accuracy in forecasting building energy loads.However,they usually lack interpretability and do not incorporate domain knowledge,making it difficult for users to trust their predictions in practical applications.One important and interesting question remains unanswered:is it possible to use intrinsically interpretable models to achieve accuracy comparable to that of black-box models?With an aim of answering this question,this study proposes an intrinsically interpretable machine learning-based method to forecast building energy loads.It creatively combines two intrinsically interpretable machine learning algorithms:clustering decision trees and adaptive multiple linear regression.Clustering decision trees aim to automatically identify various building operation conditions,allowing for the training of multiple models tailored to each condition.It can reduce the complexity of model training data,leading to higher accuracy.Adaptive multiple linear regression is an improved regression algorithm tailored to building energy load prediction.It can adaptively modify regression coefficients according to building operations,enhancing the non-linear fitting capability of multiple linear regression.The proposed method is evaluated utilizing the operational data from an office building.The results indicate that the proposed method exhibits comparable accuracy to both random forests and extreme gradient boosting.Furthermore,it shows significantly superior accuracy,with an average improvement of 10.2%,compared with some popular black-box algorithms such as artificial neural networks,support vector regression,and classification and regression trees.As for model interpretability,the proposed method reveals that historical cooling loads are the most crucial for predicting building cooling loads under most conditions.Additionally,outdoor air temperature has a significant contribution to building cooling load prediction during the daytime on weekdays in summer and transition seasons.In the future,it will be valuable to explore integrating the laws of physics into the proposed method to further enhance its interpretability.展开更多
Stand age plays a crucial role in forest biomass estimation and carbon cycle modeling.Assessing the uncertainty of stand age prediction models and identifying the key driving factors in the modeling process have becom...Stand age plays a crucial role in forest biomass estimation and carbon cycle modeling.Assessing the uncertainty of stand age prediction models and identifying the key driving factors in the modeling process have become major challenges in forestry research.In this study,we selected the Shaanxi-Gansu-Ningxia region of Northeast China as the research area and utilized multi-source datasets from the summer of 2019 to extract information on spectral,textural,climatic,water balance,and stand characteristics.By integrating the Random Forest(RF)model with Monte Carlo(MC)simulation,we constructed six regression models based on different combina-tions of features and evaluated the uncertainty of each model.Furthermore,we investigated the driving factors influencing stand age modeling by analyzing the effects of different types of features on age inversion.Model performance and accuracy were assessed using the root mean square error(RMSE),mean absolute error(MAE),and the coefficient of determination(R^(2)),while the relative root mean square error(rRMSE)was employed to quantify model uncertainty.The results indicate that the scenarios with more obvious improve-ment in accuracy and effective reduction in uncertainty were Scenario 3 with the inclusion of climate and water balance information(RMSE=25.54 yr,MAE=18.03 yr,R^(2)=0.51,rRMSE=19.17%)and Scenario 5 with the inclusion of stand characterization informa-tion(RMSE=18.47 yr,MAE=13.05 yr,R^(2)=0.74,rRMSE=16.99%).Scenario 6,incorporating all feature types,achieved the highest accuracy(RMSE=17.60 yr,MAE=12.06 yr,R^(2)=0.77,rRMSE=14.19%).In this study,elevation,minimum temperature,and diameter at breast height(DBH)emerged as the key drivers of stand-age modeling.The proposed method can be used to identify drivers and to quantify uncertainty in stand-age estimation,providing a useful reference for improving model accuracy and uncertainty assessment.展开更多
Photovoltaic (PV) modules, as essential components of solar power generation systems, significantly influence unitpower generation costs.The service life of these modules directly affects these costs. Over time, the p...Photovoltaic (PV) modules, as essential components of solar power generation systems, significantly influence unitpower generation costs.The service life of these modules directly affects these costs. Over time, the performanceof PV modules gradually declines due to internal degradation and external environmental factors.This cumulativedegradation impacts the overall reliability of photovoltaic power generation. This study addresses the complexdegradation process of PV modules by developing a two-stage Wiener process model. This approach accountsfor the distinct phases of degradation resulting from module aging and environmental influences. A powerdegradation model based on the two-stage Wiener process is constructed to describe individual differences inmodule degradation processes. To estimate the model parameters, a combination of the Expectation-Maximization(EM) algorithm and the Bayesian method is employed. Furthermore, the Schwarz Information Criterion (SIC) isutilized to identify critical change points in PV module degradation trajectories. To validate the universality andeffectiveness of the proposed method, a comparative analysis is conducted against other established life predictiontechniques for PV modules.展开更多
The acquisition,analysis,and prediction of the radar cross section(RCS)of a target have extremely important strategic significance in the military.However,the RCS values at all azimuths are hardly accessible for non-c...The acquisition,analysis,and prediction of the radar cross section(RCS)of a target have extremely important strategic significance in the military.However,the RCS values at all azimuths are hardly accessible for non-cooperative targets,due to the limitations of radar observation azimuth and detection resources.Despite their efforts to predict the azimuth-dimensional RCS value,traditional methods based on statistical theory fails to achieve the desired results because of the azimuth sensitivity of the target RCS.To address this problem,an improved neural basis expansion analysis for interpretable time series forecasting(N-BEATS)network considering the physical model prior is proposed to predict the azimuth-dimensional RCS value accurately.Concretely,physical model-based constraints are imposed on the network by constructing a scattering-center module based on the target scattering-center model.Besides,a superimposed seasonality module is involved to better capture high-frequency information,and augmenting the training set provides complementary information for learning predictions.Extensive simulations and experimental results are provided to validate the effectiveness of the proposed method.展开更多
Accurate prediction of coal and gas outburst(CGO)hazards is paramount in gas disaster prevention and control.This paper endeavors to overcome the constraints posed by traditional prediction indexes when dealing with C...Accurate prediction of coal and gas outburst(CGO)hazards is paramount in gas disaster prevention and control.This paper endeavors to overcome the constraints posed by traditional prediction indexes when dealing with CGO incidents under low gas pressure conditions.In pursuit of this objective,we have studied and established a mechanical model of the working face under abnormal stress and the excitation energy conditions of CGO,and proposed a method for predicting the risk of CGO under abnormal stress.On site application verification shows that when a strong outburst hazard level prediction is issued,there is a high possibility of outburst disasters occurring.In one of the three locations where we predicted strong outburst hazards,a small outburst occurred,and the accuracy of the prediction was higher than the traditional drilling cuttings index S and drilling cuttings gas desorption index q.Finally,we discuss the mechanism of CGO under the action of stress anomalies.Based on the analysis of stress distribution changes and energy accumulation characteristics of coal under abnormal stress,this article believes that the increase in outburst risk caused by high stress abnormal gradient is mainly due to two reasons:(1)The high stress abnormal gradient leads to an increase in the plastic zone of the coal seam.After the working face advances,it indirectly leads to an increase in the gas expansion energy that can be released from the coal seam before reaching a new stress equilibrium.(2)Abnormal stress leads to increased peak stress of coal body in front of working face.When coal body in elastic area transforms to plastic area,its failure speed is accelerated,which induces accelerated gas desorption and aggravates the risk of outburst.展开更多
The mathematical method of ZTD(zenith tropospheric delay)spatial prediction is important for precise ZTD derivation and real-time precise point positioning(PPP)augmentation.This paper analyses the performance of the p...The mathematical method of ZTD(zenith tropospheric delay)spatial prediction is important for precise ZTD derivation and real-time precise point positioning(PPP)augmentation.This paper analyses the performance of the popular optimal function coefficient(OFC),sphere cap harmonic analysis(SCHA),kriging and inverse distance weighting(IDW)interpolation in ZTD spatial prediction and Beidou satellite navigation system(BDS)-PPP augmentation over China.For ZTD spatial prediction,the average time consumption of the OFC,kriging,and IDW methods is less than 0.1 s,which is significantly better than that of the SCHA method(63.157 s).The overall ZTD precision of the OFC is 3.44 cm,which outperforms those of the SCHA(9.65 cm),Kriging(10.6 cm),and IDW(11.8 cm)methods.We confirmed that the low performance of kriging and IDW is caused by their weakness in modelling ZTD variation in the vertical direction.To mitigate such deficiencies,an elevation normalization factor(ENF)is introduced into the kriging and IDW models(kriging-ENF and IDW-ENF).The overall ZTD spatial prediction accuracies of IDW-ENF and kriging-ENF are 2.80 cm and 2.01 cm,respectively,which are both superior to those of the OFC and the widely used empirical model GPT3(4.92 cm).For BDS-PPP enhancement,the ZTD provided by the kriging-ENF,IDW-ENF and OFC as prior constraints can effectively reduce the convergence time.Compared with unconstrained BDS-PPP,our proposed kriging-ENF outperforms IDW-ENF and OFC by reducing the horizontal and vertical convergence times by approximately 13.2%and 5.8%in Ningxia and 30.4%and 7.84%in Guangdong,respectively.These results indicate that kriging-ENF is a promising method for ZTD spatial prediction and BDS-PPP enhancement over China.展开更多
Wellbore instability is one of the significant challenges in the drilling engineering and during the development of carbonate reservoirs,especially with open-hole completion.The problems of wellbore instability such a...Wellbore instability is one of the significant challenges in the drilling engineering and during the development of carbonate reservoirs,especially with open-hole completion.The problems of wellbore instability such as downhole collapse and silt deposit in the fractured carbonate reservoir of Tarim Basin(Ordovician)are severe.Solid destabilization and production(SDP)was proposed to describe this engineering problem of carbonate reservoirs.To clarify the mechanism and mitigate potential borehole instability problems,we conducted particle size distribution(PSD)analysis,X-ray diffraction(XRD)analysis,triaxial compression tests,and micro-scale sand production tests based on data analysis.We found that the rock fragments and silt in the wellbore came from two sources:one from the wellbore collapse in the upper unplugged layers and the other from the production of sand particles carried by the fluid in the productive layers.Based on the experimental study,a novel method combining a geomechanical model and microscopic sand production model was proposed to predict wellbore instability and analyze its influencing factors.The critical condition and failure zone predicted by the prediction model fit well with the field observations.According to the prediction results,the management and prevention measures of wellbore instability in carbonate reservoirs were proposed.It is suggested to optimize the well track in new drilling wells while upgrading the production system in old wells.This study is of great guiding significance for the optimization of carbonate solid control and it improves the understanding of the sand production problems in carbonate reservoirs.展开更多
Accurate and efficient prediction of the distribution of surface loads on buildings subjected to explosive effects is crucial for rapidly calculating structural dynamic responses,establishing effective protective meas...Accurate and efficient prediction of the distribution of surface loads on buildings subjected to explosive effects is crucial for rapidly calculating structural dynamic responses,establishing effective protective measures,and designing civil defense engineering solutions.Current state-of-the-art methods face several issues:Experimental research is difficult and costly to implement,theoretical research is limited to simple geometries and lacks precision,and direct simulations require substantial computational resources.To address these challenges,this paper presents a data-driven method for predicting blast loads on building surfaces.This approach increases both the accuracy and computational efficiency of load predictions when the geometry of the building changes while the explosive yield remains constant,significantly improving its applicability in complex scenarios.This study introduces an innovative encoder-decoder graph neural network model named BlastGraphNet,which uses a message-passing mechanism to predict the overpressure and impulse load distributions on buildings with conventional and complex geometries during explosive events.The model also facilitates related downstream applications,such as damage mode identification and rapid assessment of virtual city explosions.The calculation results indicate that the prediction error of the model for conventional building tests is less than 2%,and its inference speed is 3-4 orders of magnitude faster than that of state-of-the-art numerical methods.In extreme test cases involving buildings with complex geometries and building clusters,the method achieved high accuracy and excellent generalizability.The strong adaptability and generalizability of BlastGraphNet confirm that this novel method enables precise real-time prediction of blast loads and provides a new paradigm for damage assessment in protective engineering.展开更多
Stock price prediction is a typical complex time series prediction problem characterized by dynamics,nonlinearity,and complexity.This paper introduces a generative adversarial network model that incorporates an attent...Stock price prediction is a typical complex time series prediction problem characterized by dynamics,nonlinearity,and complexity.This paper introduces a generative adversarial network model that incorporates an attention mechanism(GAN-LSTM-Attention)to improve the accuracy of stock price prediction.Firstly,the generator of this model combines the Long and Short-Term Memory Network(LSTM),the Attention Mechanism and,the Fully-Connected Layer,focusing on generating the predicted stock price.The discriminator combines the Convolutional Neural Network(CNN)and the Fully-Connected Layer to discriminate between real stock prices and generated stock prices.Secondly,to evaluate the practical application ability and generalization ability of the GAN-LSTM-Attention model,four representative stocks in the United States of America(USA)stock market,namely,Standard&Poor’s 500 Index stock,Apple Incorporatedstock,AdvancedMicroDevices Incorporatedstock,and Google Incorporated stock were selected for prediction experiments,and the prediction performance was comprehensively evaluated by using the three evaluation metrics,namely,mean absolute error(MAE),root mean square error(RMSE),and coefficient of determination(R2).Finally,the specific effects of the attention mechanism,convolutional layer,and fully-connected layer on the prediction performance of the model are systematically analyzed through ablation study.The results of experiment show that the GAN-LSTM-Attention model exhibits excellent performance and robustness in stock price prediction.展开更多
Seasonal prediction of summer rainfall in China plays a crucial role in decision-making,environmental protection,and socio-economic development,while it currently has a low prediction skill.We developed a deep learnin...Seasonal prediction of summer rainfall in China plays a crucial role in decision-making,environmental protection,and socio-economic development,while it currently has a low prediction skill.We developed a deep learning-based seasonal prediction bias correction method for summer rainfall in China.Based on prediction fields from the flexible Global Ocean-Atmosphere-Land System Model finite volume version 2(FGOALS-f2),we optimized the loss function of U-Net,trained with different hyperparameters,and selected the optimum model.U-Net model can extract multi-scale feature information and preserve spatial information,making it suitable for processing meteorological data.With this endto-end model,the precipitation distribution can be obtained directly without using the traditional method of data dimensionality reduction(e.g.,Empirical Orthogonal Function),which could maximize the retention of spatio-temporal information of the input data.Optimization of the loss function enhances the prediction results and mitigates model overfitting.The independent prediction shows a significant skill improvement measured by the anomalous correlation coefficient score.The skill has an average value of 0.679 in China(0°–63°N,73°–133°E)and 0.691 in the region of the Chinese mainland,which significantly improves the dynamical prediction skill by 1357%and 4836%.This study suggests that the deep learning(U-Net)-based seasonal prediction bias correction method is a promising approach for improving rainfall prediction of the dynamical model.展开更多
Landslide susceptibility prediction(LSP)is significantly affected by the uncertainty issue of landslide related conditioning factor selection.However,most of literature only performs comparative studies on a certain c...Landslide susceptibility prediction(LSP)is significantly affected by the uncertainty issue of landslide related conditioning factor selection.However,most of literature only performs comparative studies on a certain conditioning factor selection method rather than systematically study this uncertainty issue.Targeted,this study aims to systematically explore the influence rules of various commonly used conditioning factor selection methods on LSP,and on this basis to innovatively propose a principle with universal application for optimal selection of conditioning factors.An'yuan County in southern China is taken as example considering 431 landslides and 29 types of conditioning factors.Five commonly used factor selection methods,namely,the correlation analysis(CA),linear regression(LR),principal component analysis(PCA),rough set(RS)and artificial neural network(ANN),are applied to select the optimal factor combinations from the original 29 conditioning factors.The factor selection results are then used as inputs of four types of common machine learning models to construct 20 types of combined models,such as CA-multilayer perceptron,CA-random forest.Additionally,multifactor-based multilayer perceptron random forest models that selecting conditioning factors based on the proposed principle of“accurate data,rich types,clear significance,feasible operation and avoiding duplication”are constructed for comparisons.Finally,the LSP uncertainties are evaluated by the accuracy,susceptibility index distribution,etc.Results show that:(1)multifactor-based models have generally higher LSP performance and lower uncertainties than those of factors selection-based models;(2)Influence degree of different machine learning on LSP accuracy is greater than that of different factor selection methods.Conclusively,the above commonly used conditioning factor selection methods are not ideal for improving LSP performance and may complicate the LSP processes.In contrast,a satisfied combination of conditioning factors can be constructed according to the proposed principle.展开更多
Currently,regression prediction methods based on logging data is one of the main methods for analyzing gas content of coal seams.However,the complexity of logging parameters for deep coal seams and the scarcity of mea...Currently,regression prediction methods based on logging data is one of the main methods for analyzing gas content of coal seams.However,the complexity of logging parameters for deep coal seams and the scarcity of measured gas content data signifcantly afects the accuracy and generalizability of data regression models.Accurately predicting the gas content of coal seams under small-sample condition become a difcult point in deep coalbed methane(CBM)exploration.The ModelAgnostic Meta-Learning(MAML)and Support Vector Regression(SVR)algorithms are among the few suitable for smallsample learning,exhibiting strong adaptability under limited sample conditions.In this study,logging parameters are used as input variables to construct MAML and SVR models,and their performance in predicting gas content of deep coal seams across diferent regions and layers is compared.The results demonstrate that the MAML algorithm efectively addresses the complex relationships between gas content of deep coal seam and logging parameters.The prediction errors for test dataset and new samples are merely 3.61%and 4.52%respectively,indicating exceptional adaptability,robust generalization capability,and stable model performance.In contrast,the dependency of SVR model on input parameters restricts its accuracy and generalizability in predicting gas content in deep coal seams with varying geological conditions.Although achieving a test dataset error of 4.71%,the SVR model demonstrates substantially degraded performance when applied to novel samples,with prediction errors escalating to 12.46%.Therefore,the MAML model is selected to predict gas content in the unknown areas of the Baijiahai region.The prediction results reveal that the gas content of coal seams in the Xishanyao formation(J2x)ranges from 1.32 m^(3)/t to 16.11 m^(3)/t,while that in the Badaowan Formation(J1b)varies between 1.73 m^(3)/t and 11.27 m^(3)/t.Notably,the gas enrichment areas are predominantly distributed in well blocks adjacent to fault systems,such as wells C31 and BJ8,etc.,which align with the favorable geological conditions for deep CBM accumulation in the Baijiahai region.These spatial distribution patterns not only corroborate existing geological insights but also further validate the reliability of the MAML model in predicting gas content within deep coal seams.展开更多
Porosity is an important attribute for evaluating the petrophysical properties of reservoirs, and has guiding significance for the exploration and development of oil and gas. The seismic inversion is a key method for ...Porosity is an important attribute for evaluating the petrophysical properties of reservoirs, and has guiding significance for the exploration and development of oil and gas. The seismic inversion is a key method for comprehensively obtaining the porosity. Deep learning methods provide an intelligent approach to suppress the ambiguity of the conventional inversion method. However, under the trace-bytrace inversion strategy, there is a lack of constraints from geological structural information, resulting in poor lateral continuity of prediction results. In addition, the heterogeneity and the sedimentary variability of subsurface media also lead to uncertainty in intelligent prediction. To achieve fine prediction of porosity, we consider the lateral continuity and variability and propose an improved structural modeling deep learning porosity prediction method. First, we combine well data, waveform attributes, and structural information as constraints to model geophysical parameters, constructing a high-quality training dataset with sedimentary facies-controlled significance. Subsequently, we introduce a gated axial attention mechanism to enhance the features of dataset and design a bidirectional closed-loop network system constrained by inversion and forward processes. The constraint coefficient is adaptively adjusted by the petrophysical information contained between the porosity and impedance in the study area. We demonstrate the effectiveness of the adaptive coefficient through numerical experiments.Finally, we compare the performance differences between the proposed method and conventional deep learning methods using data from two study areas. The proposed method achieves better consistency with the logging porosity, demonstrating the superiority of the proposed method.展开更多
To address prediction errors and limited information extraction in machine learning(ML)-based interval prediction,a hybrid model was proposed for interval estimation and failure assessment of step-like landslides unde...To address prediction errors and limited information extraction in machine learning(ML)-based interval prediction,a hybrid model was proposed for interval estimation and failure assessment of step-like landslides under uncertainty.The model decomposed displacements into trend and periodic components via Variational Mode Decomposition(VMD)and K-shape clustering.The Residual and Moving Block Bootstrap methods were used to generate pseudo datasets.Polynomial regressionwas adopted for trend forecasting,whereas the Dense Convolutional Network(DenseNet)and Long Short-Term Memory(LSTM)networks were employed for periodic displacement prediction.An Extreme Learning Machine(ELM)was used to estimate the noise variance,enabling the construction of Prediction Intervals(PIs)and quantificationof displacement uncertainty.Failure probabilities(Pf)were derived from PIs using an improved tangential angle criterion and reliability analysis.The model was validated on three step-like landslides in the Three Gorges Reservoir Area,achieving stability assessment accuracies of 99.88%(XD01),99.93%(ZG93),99.89%(ZG118),and 100%for ZG110 and ZG111 across the Baishuihe and Bazimen landslides.For the Shuping landslide,the predictions aligned with fieldobservations before and after the 2014–2015 remediation,with P_(f)remaining near zero post-2015 except for occasional peaks.The model outperformed conventional ML approaches by yielding narrower PIs.At XD01 with 90%PI nominal confidencelevel(PINC),the coverage width-based criterion(CWC)and PI average width(PIAW)were 3.38 mm.The mean values of the PIs exhibited high accuracy,with a Mean Absolute Error(MAE)of 0.28 mm and Root Mean Square Error(RMSE)of 0.39 mm.These results demonstrate the robustness of the proposed model in improving landslide risk assessment and decision-making under uncertainty.展开更多
基金supported by the National NaturalScience Foundation of China(Nos.52301029 and 52274359)the Fundamental Research Funds for the CentralUniversities,China(No.06500165)+2 种基金the Guangdong Basicand Applied Basic Research Foundation,China(No.2022A1515140006)Young Elite Scientists Sponsorship Program by CAST(No.2023QNRC001)Beijing Young Elite Scientists Sponsorship Program by BMES,China.
文摘In the era of materials genome engineering,data-driven machine learning has become a powerful tool for accelerating the re-search and development of metallic materials.However,the predictive accuracy and generalization ability of traditional machine learning models are often limited by the scarcity and heterogeneity of available data,especially in small-sample scenarios.To address these chal-lenges,transfer learning has emerged as an effective strategy to leverage knowledge from related domains,thereby enhancing model per-formance with limited target data.This review systematically summarizes the fundamental concepts,methodologies,and representative applications of transfer learning in the prediction of metallic materials'properties.Transfer learning can be categorized into feature-based,instance-based,parameter-based,and knowledge-based methods.This work discusses their respective mechanisms,advantages,and limit-ations.Case studies demonstrate that transfer learning can significantly improve prediction accuracy,data efficiency,and model inter-pretability in tasks such as mechanical property prediction and alloy design.Furthermore,this work highlights emerging trends including hybrid,multi-task,meta,and adaptive transfer learning,which further expand the applicability of these techniques.Finally,this work out-lines future research directions,emphasizing the need for data standardization,algorithmic innovation,multimodal data fusion,and the in-tegration of physical principles to achieve robust,interpretable,and generalizable models.The perspectives presented aim to advance the intelligent design and discovery of metallic materials,promoting efficient knowledge transfer and collaborative innovation in materials science.
基金supported in part by the National Key RD Program of China under Grant 2023YFB4204400,and in part by the National Natural Science Foundation of China under Grant 52125705.
文摘In practical microgrids,current saturation of inverters and power interaction coupling of different forms of DERs complicate the system's transient behaviors.Existing methods of online transient stability prediction(TSP)are suitable for power systems consisting of homogeneous distributed energy resources(DERs),thus showing limited accuracy for stability prediction of microgrids.This paper develops a deep-learning-based TSP method for accurate online prediction of microgrids consisting of diverse forms of DERs under current saturation.First,a general key input feature selection method for microgrid TSP is systematically designed to ensure prediction accuracy.It is derived from a comprehensive mechanism analysis of the influence of DER's intrinsic and interaction characteristics under current saturation.Besides,impacts of load fluctuation and fault change are also considered to improve robust prediction performance.Second,to further improve prediction accuracy,an online TSP model based on deep learning is developed by effectively using the powerful nonlinear mapping capability of the deep belief network(DBN).Then,by combining feature selection method and deep-learning-based TSP model,an online TSP method is derived.Test results show the proposed method greatly improves accuracy of microgrid TSP under complex operating conditions.Furthermore,the method effectively avoids feature redundancy and the curse of dimensionality.Numbers of input features are independent of the scale of microgrids.
基金supported by the CRRC Original Technology TenYear Cultivation Program(Grant No.2022CYY007)。
文摘Deep learning and fatigue life prediction remain focal research areas in rail vehicle engineering.This study addresses the vibration fatigue of wheelset lifting lug in Chengdu Metro Line 1 bogies,aiming to develop a fatigue life prediction method for critical bogie components using deep learning models and measured track load spectra.Extensive field tests on Chengdu Metro Line 1 were conducted to acquire acceleration and stress response data of the wheelset lifting lug,generating training samples for the neural network system.Component stress responses were calculated via time-domain track acceleration and validated against in-situ stress measurements.Results show that neural network-fitted dynamic stress values exhibit excellent consistency with measured data,with errors constrained within 5%.This study validates the proposed small-sample deep learning approach as an effective and accurate solution for fatigue life prediction of critical bogie components under operational load conditions.
基金Vicerrectoría de Inves-tigaciones of the UMNG for the financial support of the IMP-ING-3743 Project.
文摘This study presents a hybrid methodology for predicting building collapses within the Intelligent Circular Resilience(ICR)framework.This uses a supervised Machine Learning(ML)approach,earthquake damage re-ports,and the Simplified Resilience Index(SRI),derived from existing earthquake damage models(EDM)-based on fragility and vulnerability functions-used in the probabilistic seismic risk assessment(PSRA).A curated building damage database comprising 89 structures(71 collapsed and 18 non-collapsed)from ten countries affected by major earthquakes(Mw 6.1-8.1,epicentral distances of 3-125 km,and PGA values ranging from 0.14 g to 0.82 g)was developed,including attributes related to exposure:occupancy,main structural material,number of stories,construction year,and hazard:magnitude,epicentral distance,intensity measures(Peak-ground acceleration,PGA,and elastic spectral acceleration).The dataset includes events such as the 2017 Puebla-Morelos earthquake(Mw 7.1,Mexico),the 1999 Kocaeli earthquake(Mw 7.6,Turkey),and the 2011 Christchurch earthquake(Mw 6.1,New Zealand),among others.Likewise,dependent attributes such as time elapsed and SRI(under 120-,180-,and 365-day recovery scenarios)were calculated using 2-EDMs.Eight Random Forest models were trained and tested for collapse and non-collapse classification using combinations of independent and dependent attributes.The results indicate that models incorporating exposure-related varia-bles-such as structural material,number of stories,construction year,and occupancy-alongside the SRI significantly improve collapse classification performance,achieving recall and F1 scores above 95%.Notably,many collapsed buildings exhibited low intensities(PGA≤0.25 g),emphasizing the influence of local site effects-particularly in Mexico City.The findings demonstrate that incorporating SRI enhances the reliability of collapse prediction and supports its use as an interpretable resilience proxy during early ICR stages.This hybrid methodology bridges empirical data,traditional PSRA models,and ML techniques,contributing to more accurate and scalable post-earthquake resilience assessments.
基金supported by the Tianjin Manufacturing High Quality Development Special Foundation(No.20232185)the Roycom Foundation(No.70306901).
文摘Hard disk drives(HDDs)serve as the primary storage devices in modern data centers.Once a failure occurs,it often leads to severe data loss,significantly degrading the reliability of storage systems.Numerous studies have proposed machine learning-based HDD failure prediction models.However,the Self-Monitoring,Analysis,and Reporting Technology(SMART)attributes differ across HDD manufacturers.We define hard drives of the same brand and model as homogeneous HDD groups,and those from different brands or models as heterogeneous HDD groups.In practical engineering scenarios,a data center is often composed of a heterogeneous population of HDDs,spanning multiple vendors and models.Existing research predominantly focuses on homogeneous datasets,ignoring the model’s generalization capability across heterogeneous HDDs.As a result,HDD models with limited samples often suffer from poor training effectiveness and prediction performance.To address this issue,we investigate generalizable SMART predictors across heterogeneous HDD groups.By extracting time-series features within a fixed sliding time window,we propose a Heterogeneous Disk Failure Prediction Method based on Time Series Features(HDFPM)framework.This method is adaptable to HDD models with limited sample sizes,thereby enhancing its applicability and robustness across diverse drive populations.Experimental results show that the proposed model achieves an F1-score of 0.9518 when applied to two different Seagate HDD models,while maintaining the False Positive Rate(FPR)below 1%.After incorporating the Complexity-Ratio Dynamic Time Warping(CDTW)based feature enhancement method,the best prediction model achieves a True Positive Rate(TPR)of up to 0.93 between the two models.For next-day failure prediction across various Seagate models,the model achieves an F1-score of up to 0.8792.Moreover,the experimental results also show that within the same brand,the higher the proportion of shared SMART attributes across different models,the better the prediction performance.In addition,HDFPMdemonstrates the best stability andmost significant performance in heterogeneous environments.
基金Supported by the China National Science and Technology Major Project(2025ZD1407000)PetroChina Science and Technology Major Project(2023ZZ22)。
文摘To address the challenges of complex fluvial sandbody distribution and difficult remaining oil recovery in mature continental oilfields,this study focuses on key issues in reservoir identification such as ambiguous narrow-channel boundaries and subdivision of multi-stage superimposed sandbodies.Taking the Upper Cretaceous continental sandstone in the Sazhong Oilfield of the Daqing Placanticline as an example,a technical system integrating OVT high-resolution processing,multi-attribute fusion,and varible-scale inversion was developed to establish a complete workflow from seismic processing to reservoir prediction and remaining oil recovery.The following results are obtained.First,the Offset Vector Tile(OVT)seismic processing technology is extended,for the first time,from fracture imaging to sandbody prediction,in order to address the weak seismic responses from boundaries of narrow and thin sandbodies.A geology-oriented OVT partitioning method is developed to significantly improve the imaging accuracy,enabling identification of channel sandbodies as narrow as 50 m.Second,an amplitude-coherence dual-attribute fusion method is proposed for predicting narrow channel boundaries between wells.Constrained by a sedimentary unit-level sequence chronostratigraphic framework,this method accurately delineates 800-2000 m long subaqueous distributary channels with bifurcation-convergence features.Third,considering the superimposition of multi-stage channels,a three-level variable-scale stratigraphic model(sandstone groups,sublayers,sedimentary units)is constructed to overcome single-scale modeling limitations,successfully characterizing key sedimentary features like meandering river“cut-offs”through 3D seismic inversion.Based on these advances,a direct link between seismic prediction and remaining oil recovery is established.The horizontal wells deployed using narrow-channel predictions encountered oil-bearing sandstones in the horizontal section by 97%,and achieved initial daily production of 12.5 t per well.Precise identification of individual channel boundaries within 17 composite sandbodies guided recovery processes in 135 wells,yielding an average daily increase of 2.8 t per well and a cumulative increase of 13.6×10^(4)t.
基金supported by the National Natural Science Foundation of China(No.52161135202).
文摘Black-box models have demonstrated remarkable accuracy in forecasting building energy loads.However,they usually lack interpretability and do not incorporate domain knowledge,making it difficult for users to trust their predictions in practical applications.One important and interesting question remains unanswered:is it possible to use intrinsically interpretable models to achieve accuracy comparable to that of black-box models?With an aim of answering this question,this study proposes an intrinsically interpretable machine learning-based method to forecast building energy loads.It creatively combines two intrinsically interpretable machine learning algorithms:clustering decision trees and adaptive multiple linear regression.Clustering decision trees aim to automatically identify various building operation conditions,allowing for the training of multiple models tailored to each condition.It can reduce the complexity of model training data,leading to higher accuracy.Adaptive multiple linear regression is an improved regression algorithm tailored to building energy load prediction.It can adaptively modify regression coefficients according to building operations,enhancing the non-linear fitting capability of multiple linear regression.The proposed method is evaluated utilizing the operational data from an office building.The results indicate that the proposed method exhibits comparable accuracy to both random forests and extreme gradient boosting.Furthermore,it shows significantly superior accuracy,with an average improvement of 10.2%,compared with some popular black-box algorithms such as artificial neural networks,support vector regression,and classification and regression trees.As for model interpretability,the proposed method reveals that historical cooling loads are the most crucial for predicting building cooling loads under most conditions.Additionally,outdoor air temperature has a significant contribution to building cooling load prediction during the daytime on weekdays in summer and transition seasons.In the future,it will be valuable to explore integrating the laws of physics into the proposed method to further enhance its interpretability.
基金Under the auspices of the Natural Science Foundation of China(No.32371875,32001249)。
文摘Stand age plays a crucial role in forest biomass estimation and carbon cycle modeling.Assessing the uncertainty of stand age prediction models and identifying the key driving factors in the modeling process have become major challenges in forestry research.In this study,we selected the Shaanxi-Gansu-Ningxia region of Northeast China as the research area and utilized multi-source datasets from the summer of 2019 to extract information on spectral,textural,climatic,water balance,and stand characteristics.By integrating the Random Forest(RF)model with Monte Carlo(MC)simulation,we constructed six regression models based on different combina-tions of features and evaluated the uncertainty of each model.Furthermore,we investigated the driving factors influencing stand age modeling by analyzing the effects of different types of features on age inversion.Model performance and accuracy were assessed using the root mean square error(RMSE),mean absolute error(MAE),and the coefficient of determination(R^(2)),while the relative root mean square error(rRMSE)was employed to quantify model uncertainty.The results indicate that the scenarios with more obvious improve-ment in accuracy and effective reduction in uncertainty were Scenario 3 with the inclusion of climate and water balance information(RMSE=25.54 yr,MAE=18.03 yr,R^(2)=0.51,rRMSE=19.17%)and Scenario 5 with the inclusion of stand characterization informa-tion(RMSE=18.47 yr,MAE=13.05 yr,R^(2)=0.74,rRMSE=16.99%).Scenario 6,incorporating all feature types,achieved the highest accuracy(RMSE=17.60 yr,MAE=12.06 yr,R^(2)=0.77,rRMSE=14.19%).In this study,elevation,minimum temperature,and diameter at breast height(DBH)emerged as the key drivers of stand-age modeling.The proposed method can be used to identify drivers and to quantify uncertainty in stand-age estimation,providing a useful reference for improving model accuracy and uncertainty assessment.
基金supported by the National Natural Science Foundation of China(51767017)the Basic Research Innovation Group Project of Gansu Province(18JR3RA133)the Industrial Support and Guidance Project of Universities in Gansu Province(2022CYZC-22).
文摘Photovoltaic (PV) modules, as essential components of solar power generation systems, significantly influence unitpower generation costs.The service life of these modules directly affects these costs. Over time, the performanceof PV modules gradually declines due to internal degradation and external environmental factors.This cumulativedegradation impacts the overall reliability of photovoltaic power generation. This study addresses the complexdegradation process of PV modules by developing a two-stage Wiener process model. This approach accountsfor the distinct phases of degradation resulting from module aging and environmental influences. A powerdegradation model based on the two-stage Wiener process is constructed to describe individual differences inmodule degradation processes. To estimate the model parameters, a combination of the Expectation-Maximization(EM) algorithm and the Bayesian method is employed. Furthermore, the Schwarz Information Criterion (SIC) isutilized to identify critical change points in PV module degradation trajectories. To validate the universality andeffectiveness of the proposed method, a comparative analysis is conducted against other established life predictiontechniques for PV modules.
基金National Natural Science Foundation of China(61921001,62201588,62022091)。
文摘The acquisition,analysis,and prediction of the radar cross section(RCS)of a target have extremely important strategic significance in the military.However,the RCS values at all azimuths are hardly accessible for non-cooperative targets,due to the limitations of radar observation azimuth and detection resources.Despite their efforts to predict the azimuth-dimensional RCS value,traditional methods based on statistical theory fails to achieve the desired results because of the azimuth sensitivity of the target RCS.To address this problem,an improved neural basis expansion analysis for interpretable time series forecasting(N-BEATS)network considering the physical model prior is proposed to predict the azimuth-dimensional RCS value accurately.Concretely,physical model-based constraints are imposed on the network by constructing a scattering-center module based on the target scattering-center model.Besides,a superimposed seasonality module is involved to better capture high-frequency information,and augmenting the training set provides complementary information for learning predictions.Extensive simulations and experimental results are provided to validate the effectiveness of the proposed method.
基金supported by the National Natural Science Foundation of China(52174162)the Fundamental Research Funds for the Central Universities(FRF-TP-20-002A3).
文摘Accurate prediction of coal and gas outburst(CGO)hazards is paramount in gas disaster prevention and control.This paper endeavors to overcome the constraints posed by traditional prediction indexes when dealing with CGO incidents under low gas pressure conditions.In pursuit of this objective,we have studied and established a mechanical model of the working face under abnormal stress and the excitation energy conditions of CGO,and proposed a method for predicting the risk of CGO under abnormal stress.On site application verification shows that when a strong outburst hazard level prediction is issued,there is a high possibility of outburst disasters occurring.In one of the three locations where we predicted strong outburst hazards,a small outburst occurred,and the accuracy of the prediction was higher than the traditional drilling cuttings index S and drilling cuttings gas desorption index q.Finally,we discuss the mechanism of CGO under the action of stress anomalies.Based on the analysis of stress distribution changes and energy accumulation characteristics of coal under abnormal stress,this article believes that the increase in outburst risk caused by high stress abnormal gradient is mainly due to two reasons:(1)The high stress abnormal gradient leads to an increase in the plastic zone of the coal seam.After the working face advances,it indirectly leads to an increase in the gas expansion energy that can be released from the coal seam before reaching a new stress equilibrium.(2)Abnormal stress leads to increased peak stress of coal body in front of working face.When coal body in elastic area transforms to plastic area,its failure speed is accelerated,which induces accelerated gas desorption and aggravates the risk of outburst.
基金co-supported by the National Nature Science Foundation of China(No.12303071)the Shanghai Science and Technology Plan Project,China(No.23YF1455500)+1 种基金the China Postdoctoral Science Foundation(No.2023M743653)Ministry of Industry and Information Technology of China through the High Precision Timing Service Project(No.TC220A04A-80)。
文摘The mathematical method of ZTD(zenith tropospheric delay)spatial prediction is important for precise ZTD derivation and real-time precise point positioning(PPP)augmentation.This paper analyses the performance of the popular optimal function coefficient(OFC),sphere cap harmonic analysis(SCHA),kriging and inverse distance weighting(IDW)interpolation in ZTD spatial prediction and Beidou satellite navigation system(BDS)-PPP augmentation over China.For ZTD spatial prediction,the average time consumption of the OFC,kriging,and IDW methods is less than 0.1 s,which is significantly better than that of the SCHA method(63.157 s).The overall ZTD precision of the OFC is 3.44 cm,which outperforms those of the SCHA(9.65 cm),Kriging(10.6 cm),and IDW(11.8 cm)methods.We confirmed that the low performance of kriging and IDW is caused by their weakness in modelling ZTD variation in the vertical direction.To mitigate such deficiencies,an elevation normalization factor(ENF)is introduced into the kriging and IDW models(kriging-ENF and IDW-ENF).The overall ZTD spatial prediction accuracies of IDW-ENF and kriging-ENF are 2.80 cm and 2.01 cm,respectively,which are both superior to those of the OFC and the widely used empirical model GPT3(4.92 cm).For BDS-PPP enhancement,the ZTD provided by the kriging-ENF,IDW-ENF and OFC as prior constraints can effectively reduce the convergence time.Compared with unconstrained BDS-PPP,our proposed kriging-ENF outperforms IDW-ENF and OFC by reducing the horizontal and vertical convergence times by approximately 13.2%and 5.8%in Ningxia and 30.4%and 7.84%in Guangdong,respectively.These results indicate that kriging-ENF is a promising method for ZTD spatial prediction and BDS-PPP enhancement over China.
基金financially supported by the National Natural Science Foundation of China(Grant No.52074331).
文摘Wellbore instability is one of the significant challenges in the drilling engineering and during the development of carbonate reservoirs,especially with open-hole completion.The problems of wellbore instability such as downhole collapse and silt deposit in the fractured carbonate reservoir of Tarim Basin(Ordovician)are severe.Solid destabilization and production(SDP)was proposed to describe this engineering problem of carbonate reservoirs.To clarify the mechanism and mitigate potential borehole instability problems,we conducted particle size distribution(PSD)analysis,X-ray diffraction(XRD)analysis,triaxial compression tests,and micro-scale sand production tests based on data analysis.We found that the rock fragments and silt in the wellbore came from two sources:one from the wellbore collapse in the upper unplugged layers and the other from the production of sand particles carried by the fluid in the productive layers.Based on the experimental study,a novel method combining a geomechanical model and microscopic sand production model was proposed to predict wellbore instability and analyze its influencing factors.The critical condition and failure zone predicted by the prediction model fit well with the field observations.According to the prediction results,the management and prevention measures of wellbore instability in carbonate reservoirs were proposed.It is suggested to optimize the well track in new drilling wells while upgrading the production system in old wells.This study is of great guiding significance for the optimization of carbonate solid control and it improves the understanding of the sand production problems in carbonate reservoirs.
基金supported by the National Natural Science Founion of China(U2241285).
文摘Accurate and efficient prediction of the distribution of surface loads on buildings subjected to explosive effects is crucial for rapidly calculating structural dynamic responses,establishing effective protective measures,and designing civil defense engineering solutions.Current state-of-the-art methods face several issues:Experimental research is difficult and costly to implement,theoretical research is limited to simple geometries and lacks precision,and direct simulations require substantial computational resources.To address these challenges,this paper presents a data-driven method for predicting blast loads on building surfaces.This approach increases both the accuracy and computational efficiency of load predictions when the geometry of the building changes while the explosive yield remains constant,significantly improving its applicability in complex scenarios.This study introduces an innovative encoder-decoder graph neural network model named BlastGraphNet,which uses a message-passing mechanism to predict the overpressure and impulse load distributions on buildings with conventional and complex geometries during explosive events.The model also facilitates related downstream applications,such as damage mode identification and rapid assessment of virtual city explosions.The calculation results indicate that the prediction error of the model for conventional building tests is less than 2%,and its inference speed is 3-4 orders of magnitude faster than that of state-of-the-art numerical methods.In extreme test cases involving buildings with complex geometries and building clusters,the method achieved high accuracy and excellent generalizability.The strong adaptability and generalizability of BlastGraphNet confirm that this novel method enables precise real-time prediction of blast loads and provides a new paradigm for damage assessment in protective engineering.
基金funded by the project supported by the Natural Science Foundation of Heilongjiang Provincial(Grant Number LH2023F033)the Science and Technology Innovation Talent Project of Harbin(Grant Number 2022CXRCCG006).
文摘Stock price prediction is a typical complex time series prediction problem characterized by dynamics,nonlinearity,and complexity.This paper introduces a generative adversarial network model that incorporates an attention mechanism(GAN-LSTM-Attention)to improve the accuracy of stock price prediction.Firstly,the generator of this model combines the Long and Short-Term Memory Network(LSTM),the Attention Mechanism and,the Fully-Connected Layer,focusing on generating the predicted stock price.The discriminator combines the Convolutional Neural Network(CNN)and the Fully-Connected Layer to discriminate between real stock prices and generated stock prices.Secondly,to evaluate the practical application ability and generalization ability of the GAN-LSTM-Attention model,four representative stocks in the United States of America(USA)stock market,namely,Standard&Poor’s 500 Index stock,Apple Incorporatedstock,AdvancedMicroDevices Incorporatedstock,and Google Incorporated stock were selected for prediction experiments,and the prediction performance was comprehensively evaluated by using the three evaluation metrics,namely,mean absolute error(MAE),root mean square error(RMSE),and coefficient of determination(R2).Finally,the specific effects of the attention mechanism,convolutional layer,and fully-connected layer on the prediction performance of the model are systematically analyzed through ablation study.The results of experiment show that the GAN-LSTM-Attention model exhibits excellent performance and robustness in stock price prediction.
基金Guangdong Major Project of Basic and Applied Basic Research(2020B0301030004)Postdoctoral Fellowship Program of CPSF(GZC20232598)+1 种基金China Postdoctoral Science Foundation(2024M753168)National Key Scientific and Technological Infrastructure Project“Earth System Numerical Simulation Facility”(EarthLab)。
文摘Seasonal prediction of summer rainfall in China plays a crucial role in decision-making,environmental protection,and socio-economic development,while it currently has a low prediction skill.We developed a deep learning-based seasonal prediction bias correction method for summer rainfall in China.Based on prediction fields from the flexible Global Ocean-Atmosphere-Land System Model finite volume version 2(FGOALS-f2),we optimized the loss function of U-Net,trained with different hyperparameters,and selected the optimum model.U-Net model can extract multi-scale feature information and preserve spatial information,making it suitable for processing meteorological data.With this endto-end model,the precipitation distribution can be obtained directly without using the traditional method of data dimensionality reduction(e.g.,Empirical Orthogonal Function),which could maximize the retention of spatio-temporal information of the input data.Optimization of the loss function enhances the prediction results and mitigates model overfitting.The independent prediction shows a significant skill improvement measured by the anomalous correlation coefficient score.The skill has an average value of 0.679 in China(0°–63°N,73°–133°E)and 0.691 in the region of the Chinese mainland,which significantly improves the dynamical prediction skill by 1357%and 4836%.This study suggests that the deep learning(U-Net)-based seasonal prediction bias correction method is a promising approach for improving rainfall prediction of the dynamical model.
基金funded by the Natural Science Foundation of China(Grant Nos.42377164 and 41972280)the Badong National Observation and Research Station of Geohazards(Grant No.BNORSG-202305).
文摘Landslide susceptibility prediction(LSP)is significantly affected by the uncertainty issue of landslide related conditioning factor selection.However,most of literature only performs comparative studies on a certain conditioning factor selection method rather than systematically study this uncertainty issue.Targeted,this study aims to systematically explore the influence rules of various commonly used conditioning factor selection methods on LSP,and on this basis to innovatively propose a principle with universal application for optimal selection of conditioning factors.An'yuan County in southern China is taken as example considering 431 landslides and 29 types of conditioning factors.Five commonly used factor selection methods,namely,the correlation analysis(CA),linear regression(LR),principal component analysis(PCA),rough set(RS)and artificial neural network(ANN),are applied to select the optimal factor combinations from the original 29 conditioning factors.The factor selection results are then used as inputs of four types of common machine learning models to construct 20 types of combined models,such as CA-multilayer perceptron,CA-random forest.Additionally,multifactor-based multilayer perceptron random forest models that selecting conditioning factors based on the proposed principle of“accurate data,rich types,clear significance,feasible operation and avoiding duplication”are constructed for comparisons.Finally,the LSP uncertainties are evaluated by the accuracy,susceptibility index distribution,etc.Results show that:(1)multifactor-based models have generally higher LSP performance and lower uncertainties than those of factors selection-based models;(2)Influence degree of different machine learning on LSP accuracy is greater than that of different factor selection methods.Conclusively,the above commonly used conditioning factor selection methods are not ideal for improving LSP performance and may complicate the LSP processes.In contrast,a satisfied combination of conditioning factors can be constructed according to the proposed principle.
基金supported by the National Natural Science Foundation of China(42272200)The Science and Technology Major Project of China National Petroleum Corporation(2023ZZ18-03)+1 种基金The Science and Technology Major Project of Changqing Oilfeld(2023DZZ01)The Technology project of Huaneng Group Headquarters(Medium-deep Low-Rank Coalbed Methane Resource Potential Evaluation and Key Development Technologies of Zhalainuoer Coalfeld,HNKJ23-H51).
文摘Currently,regression prediction methods based on logging data is one of the main methods for analyzing gas content of coal seams.However,the complexity of logging parameters for deep coal seams and the scarcity of measured gas content data signifcantly afects the accuracy and generalizability of data regression models.Accurately predicting the gas content of coal seams under small-sample condition become a difcult point in deep coalbed methane(CBM)exploration.The ModelAgnostic Meta-Learning(MAML)and Support Vector Regression(SVR)algorithms are among the few suitable for smallsample learning,exhibiting strong adaptability under limited sample conditions.In this study,logging parameters are used as input variables to construct MAML and SVR models,and their performance in predicting gas content of deep coal seams across diferent regions and layers is compared.The results demonstrate that the MAML algorithm efectively addresses the complex relationships between gas content of deep coal seam and logging parameters.The prediction errors for test dataset and new samples are merely 3.61%and 4.52%respectively,indicating exceptional adaptability,robust generalization capability,and stable model performance.In contrast,the dependency of SVR model on input parameters restricts its accuracy and generalizability in predicting gas content in deep coal seams with varying geological conditions.Although achieving a test dataset error of 4.71%,the SVR model demonstrates substantially degraded performance when applied to novel samples,with prediction errors escalating to 12.46%.Therefore,the MAML model is selected to predict gas content in the unknown areas of the Baijiahai region.The prediction results reveal that the gas content of coal seams in the Xishanyao formation(J2x)ranges from 1.32 m^(3)/t to 16.11 m^(3)/t,while that in the Badaowan Formation(J1b)varies between 1.73 m^(3)/t and 11.27 m^(3)/t.Notably,the gas enrichment areas are predominantly distributed in well blocks adjacent to fault systems,such as wells C31 and BJ8,etc.,which align with the favorable geological conditions for deep CBM accumulation in the Baijiahai region.These spatial distribution patterns not only corroborate existing geological insights but also further validate the reliability of the MAML model in predicting gas content within deep coal seams.
基金the support of Research Program of Fine Exploration and Surrounding Rock Classification Technology for Deep Buried Long Tunnels Driven by Horizontal Directional Drilling and Magnetotelluric Methods Based on Deep Learning under Grant E202408010the Sichuan Science and Technology Program under Grant 2024NSFSC1984 and Grant 2024NSFSC1990。
文摘Porosity is an important attribute for evaluating the petrophysical properties of reservoirs, and has guiding significance for the exploration and development of oil and gas. The seismic inversion is a key method for comprehensively obtaining the porosity. Deep learning methods provide an intelligent approach to suppress the ambiguity of the conventional inversion method. However, under the trace-bytrace inversion strategy, there is a lack of constraints from geological structural information, resulting in poor lateral continuity of prediction results. In addition, the heterogeneity and the sedimentary variability of subsurface media also lead to uncertainty in intelligent prediction. To achieve fine prediction of porosity, we consider the lateral continuity and variability and propose an improved structural modeling deep learning porosity prediction method. First, we combine well data, waveform attributes, and structural information as constraints to model geophysical parameters, constructing a high-quality training dataset with sedimentary facies-controlled significance. Subsequently, we introduce a gated axial attention mechanism to enhance the features of dataset and design a bidirectional closed-loop network system constrained by inversion and forward processes. The constraint coefficient is adaptively adjusted by the petrophysical information contained between the porosity and impedance in the study area. We demonstrate the effectiveness of the adaptive coefficient through numerical experiments.Finally, we compare the performance differences between the proposed method and conventional deep learning methods using data from two study areas. The proposed method achieves better consistency with the logging porosity, demonstrating the superiority of the proposed method.
基金funding support from the National Science Fund for Distinguished Young Scholars(Grant No.52125904)the National Key R&D Plan(Grant No.2022YFC3004403)the National Natural Science Foundation of China(Grant No.52039008).
文摘To address prediction errors and limited information extraction in machine learning(ML)-based interval prediction,a hybrid model was proposed for interval estimation and failure assessment of step-like landslides under uncertainty.The model decomposed displacements into trend and periodic components via Variational Mode Decomposition(VMD)and K-shape clustering.The Residual and Moving Block Bootstrap methods were used to generate pseudo datasets.Polynomial regressionwas adopted for trend forecasting,whereas the Dense Convolutional Network(DenseNet)and Long Short-Term Memory(LSTM)networks were employed for periodic displacement prediction.An Extreme Learning Machine(ELM)was used to estimate the noise variance,enabling the construction of Prediction Intervals(PIs)and quantificationof displacement uncertainty.Failure probabilities(Pf)were derived from PIs using an improved tangential angle criterion and reliability analysis.The model was validated on three step-like landslides in the Three Gorges Reservoir Area,achieving stability assessment accuracies of 99.88%(XD01),99.93%(ZG93),99.89%(ZG118),and 100%for ZG110 and ZG111 across the Baishuihe and Bazimen landslides.For the Shuping landslide,the predictions aligned with fieldobservations before and after the 2014–2015 remediation,with P_(f)remaining near zero post-2015 except for occasional peaks.The model outperformed conventional ML approaches by yielding narrower PIs.At XD01 with 90%PI nominal confidencelevel(PINC),the coverage width-based criterion(CWC)and PI average width(PIAW)were 3.38 mm.The mean values of the PIs exhibited high accuracy,with a Mean Absolute Error(MAE)of 0.28 mm and Root Mean Square Error(RMSE)of 0.39 mm.These results demonstrate the robustness of the proposed model in improving landslide risk assessment and decision-making under uncertainty.
