Well logging technology has accumulated a large amount of historical data through four generations of technological development,which forms the basis of well logging big data and digital assets.However,the value of th...Well logging technology has accumulated a large amount of historical data through four generations of technological development,which forms the basis of well logging big data and digital assets.However,the value of these data has not been well stored,managed and mined.With the development of cloud computing technology,it provides a rare development opportunity for logging big data private cloud.The traditional petrophysical evaluation and interpretation model has encountered great challenges in the face of new evaluation objects.The solution research of logging big data distributed storage,processing and learning functions integrated in logging big data private cloud has not been carried out yet.To establish a distributed logging big-data private cloud platform centered on a unifi ed learning model,which achieves the distributed storage and processing of logging big data and facilitates the learning of novel knowledge patterns via the unifi ed logging learning model integrating physical simulation and data models in a large-scale functional space,thus resolving the geo-engineering evaluation problem of geothermal fi elds.Based on the research idea of“logging big data cloud platform-unifi ed logging learning model-large function space-knowledge learning&discovery-application”,the theoretical foundation of unified learning model,cloud platform architecture,data storage and learning algorithm,arithmetic power allocation and platform monitoring,platform stability,data security,etc.have been carried on analysis.The designed logging big data cloud platform realizes parallel distributed storage and processing of data and learning algorithms.The feasibility of constructing a well logging big data cloud platform based on a unifi ed learning model of physics and data is analyzed in terms of the structure,ecology,management and security of the cloud platform.The case study shows that the logging big data cloud platform has obvious technical advantages over traditional logging evaluation methods in terms of knowledge discovery method,data software and results sharing,accuracy,speed and complexity.展开更多
Electromagnetic technology used in logging while drilling(LWD) provides the resistivity distribution around a borehole within a range of several tens of meters.However,a blind zone appears in front of the drill bit wh...Electromagnetic technology used in logging while drilling(LWD) provides the resistivity distribution around a borehole within a range of several tens of meters.However,a blind zone appears in front of the drill bit when operating in high-angle wells,limiting the ability to detect formations ahead of the drill bit.Look-ahead technology addresses this issue and substantially enhances the proactive capability of geological directional drilling.In this study,we examine the detection capabilities of various component combinations of magnetic dipole antenna.Based on the sensitivity of each component to the axial information,a coaxial component is selected as a boundary indicator.We investigate the impact of various factors,such as frequency and transmitter and receiver(TR) distance,under different geological models.This study proposes 5 and 20 kHz as appropriate frequencies,and 10-14 and 12-17 m as suitable TR distance combinations.The accuracy of the numerical calculation results is verified via air-sea testing,confirming the instrument's detection capability.A test model that eliminated the influence of environmental factors and seawater depth is developed.The results have demonstrated that the tool can recognize the interface between layers up to 21.6 m ahead.It provides a validation idea for the design of new instruments as well as the validation of detection capabilities.展开更多
China,as the world’s largest coal producer and consumer,faces increasingly severe challenges from coal mine goaf areas formed through decades of intensive mining.These underground voids,resulting from exhausted resou...China,as the world’s largest coal producer and consumer,faces increasingly severe challenges from coal mine goaf areas formed through decades of intensive mining.These underground voids,resulting from exhausted resources or technical limitations,not only cause environmental issues like land subsidence and groundwater contamination but also pose critical safety risks for ongoing mining operations,including water inrushes,gas outbursts,and roof collapses.Conventional geophysical methods such as seismic surveys and electromagnetic detection demonstrate limited effectiveness in complex geological conditions due to susceptibility to electrical heterogeneity,electromagnetic interference,and interpretation ambiguities.This study presents an innovative integrated approach combining the Audio-Frequency Electrical Transillumination(AFET)method with multi-parameter borehole logging to establish a three-dimensional detection system.The AFET technique employs 0.1–10 kHz electromagnetic waves to identify electrical anomalies associated with goafs,enabling extensive horizontal scanning.This is complemented by vertical high-resolution profiling through borehole measurements of resistivity,spontaneous potential,and acoustic velocity.Field applications in Shanxi Province’s typical coal mines achieved breakthrough results:Using a grid-drilling pattern(15 m spacing,300 m depth),the method successfully detected three concealed goafs missed by conventional approaches,with spatial positioning errors under 0.5 m.Notably,it accurately identified two un-collapsed water-filled cavities.This surface-borehole synergistic approach overcomes single-method limitations,enhancing goaf detection accuracy to over 92%.The technique provides reliable technical support for safe mining practices and represents significant progress in precise detection of hidden geological hazards in Chinese coal mines,offering valuable insights for global mining geophysics.展开更多
To improve the accuracy and generalization of well logging curve reconstruction,this paper proposes an artificial intelligence large language model“Gaia”and conducts model evaluation experiments.By fine-tuning the p...To improve the accuracy and generalization of well logging curve reconstruction,this paper proposes an artificial intelligence large language model“Gaia”and conducts model evaluation experiments.By fine-tuning the pre-trained large language model,the Gaia significantly improved its ability in extracting sequential patterns and spatial features from well-log curves.Leveraging the adapter method for fine-tuning,this model required training only about 1/70 of its original parameters,greatly improving training efficiency.Comparative experiments,ablation experiments,and generalization experiments were designed and conducted using well-log data from 250 wells.In the comparative experiment,the Gaia model was benchmarked against cutting-edge small deep learning models and conventional large language models,demonstrating that the Gaia model reduced the mean absolute error(MAE)by at least 20%.In the ablation experiments,the synergistic effect of the Gaia model's multiple components was validated,with its MAE being at least 30%lower than that of single-component models.In the generalization experiments,the superior performance of the Gaia model in blind-well predictions was further confirmed.Compared to traditional models,the Gaia model is significantly superior in accuracy and generalization for logging curve reconstruction,fully showcasing the potential of large language models in the field of well-logging.This provides a new approach for future intelligent logging data processing.展开更多
Global forest cover is undergoing significant transformations due to anthropogenic activities and natural disturbances,profoundly impacting hydrological processes.However,the inherent spatial heterogeneity within wate...Global forest cover is undergoing significant transformations due to anthropogenic activities and natural disturbances,profoundly impacting hydrological processes.However,the inherent spatial heterogeneity within watersheds leads to varied hydrological responses across spatiotemporal scales,challenging comprehensive assessment of logging impacts at the watershed scale.Here,we developed multiple forest logging scenarios using the soil and water assessment tool(SWAT)model for the Le'an River watershed,a 5,837 km2 subtropical watershed in China,to quantify the hydrological effects of forest logging across different spatiotemporal scales.Our results demonstrate that increasing forest logging ratios from 1.54% to 9.25% consistently enhanced ecohydrological sensitivity.However,sensitivity varied across spatiotemporal scales,with the rainy season(15.30%-15.81%)showing higher sensitivity than annual(11.56%-12.07%)and dry season(3.38%-5.57%)periods.Additionally,the ecohydrological sensitivity of logging varied significantly across the watershed,with midstream areas exhibiting the highest sensitivity(13.13%-13.25%),followed by downstream(11.87%-11.98%)and upstream regions(9.96%-10.05%).Furthermore,the whole watershed exhibited greater hydrological resilience to logging compared to upstream areas,with attenuated runoff changes due to scale effects.Scale effects were more pronounced during dry seasons((-8.13 to -42.13)×10^(4) m^(3)·month^(-1))than in the rainy season((-11.11 to -26.65)×10^(4) m^(3)·month^(-1)).These findings advance understanding of logging impacts on hydrology across different spatiotemporal scales in subtropical regions,providing valuable insights for forest management under increasing anthropogenic activities and climate change.展开更多
Well logging curves serve as indicators of strata attribute changes and are frequently utilized for stratigraphic analysis and comparison.Deep learning,known for its robust feature extraction capabilities,has seen con...Well logging curves serve as indicators of strata attribute changes and are frequently utilized for stratigraphic analysis and comparison.Deep learning,known for its robust feature extraction capabilities,has seen continuous adoption by scholars in the realm of well logging stratigraphic correlation tasks.Nonetheless,current deep learning algorithms often struggle to accurately capture feature changes occurring at layer boundaries within the curves.Moreover,when faced with data imbalance issues,neural networks encounter challenges in accurately modeling the one-hot encoded curve stratification positions,resulting in significant deviations between predicted and actual stratification positions.Addressing these challenges,this study proposes a novel well logging curve stratigraphic comparison algorithm based on uniformly distributed soft labels.In the training phase,a label smoothing loss function is introduced to comprehensively account for the substantial loss stemming from data imbalance and to consider the similarity between diff erent layer data.Concurrently,spatial attention and channel attention mechanisms are incorporated into the shallow and deep encoder stages of U²-Net,respectively,to better focus on changes in stratification positions.During the prediction phase,an optimized confidence threshold algorithm is proposed to constrain stratification results and solve the problem of reduced prediction accuracy because of occasional layer repetition.The proposed method is applied to real-world well logging data in oil fields.Quantitative evaluation results demonstrate that within error ranges of 1,2,and 3 m,the accuracy of well logging curve stratigraphic division reaches 87.27%,92.68%,and 95.08%,respectively,thus validating the eff ectiveness of the algorithm presented in this paper.展开更多
On March 3,2024,the prototype permeability logging instrument independently developed in China successfully completed its first downhole test in Ren 91 standard well in PetroChina Huabei Oilfield.In the open hole sect...On March 3,2024,the prototype permeability logging instrument independently developed in China successfully completed its first downhole test in Ren 91 standard well in PetroChina Huabei Oilfield.In the open hole section at a depth of 3925 metres and at a temperature of 148℃,the device collected high-quality permeability logging data.This marks a key technological breakthrough from 0 to 1 in permeability logging,and lays the foundation for the next step in developing a complete set of permeability logging equipment.展开更多
Identification of reservoir types in deep carbonates has always been a great challenge due to complex logging responses caused by the heterogeneous scale and distribution of storage spaces.Traditional cross-plot analy...Identification of reservoir types in deep carbonates has always been a great challenge due to complex logging responses caused by the heterogeneous scale and distribution of storage spaces.Traditional cross-plot analysis and empirical formula methods for identifying reservoir types using geophysical logging data have high uncertainty and low efficiency,which cannot accurately reflect the nonlinear relationship between reservoir types and logging data.Recently,the kernel Fisher discriminant analysis(KFD),a kernel-based machine learning technique,attracts attention in many fields because of its strong nonlinear processing ability.However,the overall performance of KFD model may be limited as a single kernel function cannot simultaneously extrapolate and interpolate well,especially for highly complex data cases.To address this issue,in this study,a mixed kernel Fisher discriminant analysis(MKFD)model was established and applied to identify reservoir types of the deep Sinian carbonates in central Sichuan Basin,China.The MKFD model was trained and tested with 453 datasets from 7 coring wells,utilizing GR,CAL,DEN,AC,CNL and RT logs as input variables.The particle swarm optimization(PSO)was adopted for hyper-parameter optimization of MKFD model.To evaluate the model performance,prediction results of MKFD were compared with those of basic-kernel based KFD,RF and SVM models.Subsequently,the built MKFD model was applied in a blind well test,and a variable importance analysis was conducted.The comparison and blind test results demonstrated that MKFD outperformed traditional KFD,RF and SVM in the identification of reservoir types,which provided higher accuracy and stronger generalization.The MKFD can therefore be a reliable method for identifying reservoir types of deep carbonates.展开更多
Volcanic oil and gas reservoirs are generally buried deep,which leads to a high whole-well coring cost,and the degree of development and size of reservoirs are controlled by volcanic facies.Therefore,accurately identi...Volcanic oil and gas reservoirs are generally buried deep,which leads to a high whole-well coring cost,and the degree of development and size of reservoirs are controlled by volcanic facies.Therefore,accurately identifying volcanic facies by logging curves not only provides the basis of volcanic reservoir prediction but also saves costs during exploration.The Songliao Basin is a‘fault-depression superimposed’composite basin with a typical binary filling structure.Abundant types of volcanic lithologies and facies are present in the Lishu fault depression.Volcanic activity is frequent during the sedimentary period of the Huoshiling Formation.Through systematic petrographic identification of the key exploratory well(SN165C)of the Lishu fault-depression,which is a whole-well core,it is found that the Huoshiling Formation in SN165C contains four facies and six subfacies,including the volcanic conduit facies(crypto explosive breccia subfacies),explosive facies(pyroclastic flow and thermal wave base subfacies),effusive facies(upper and lower subfacies),and volcanogenic sedimentary facies(pyroclastic sedimentary subfacies).Combining core,thin section,and logging data,the authors established identification markers and petrographic chart logging phases,and also interpreted the longitudinal variation in volcanic petro-graphic response characteristics to make the charts more applicable to this area's volcanic petrographic interpretation of the Huoshiling Formation.These charts can provide a basis for the further exploration and development of volcanic oil and gas in this area.展开更多
Logging data and its interpretation results are one of the most important basic data for understanding reservoirs and oilfield development. Standardized and unified logging interpretation results play a decisive role ...Logging data and its interpretation results are one of the most important basic data for understanding reservoirs and oilfield development. Standardized and unified logging interpretation results play a decisive role in fine reservoir description and reservoir development. Aiming at the problem of the conflict between the development effect and the initial interpretation result of Yan 9 reservoir in Hujianshan area of Ordos Basin, by combining the current well production performance, logging, oil test, production test and other data, on the basis of making full use of core, coring, logging, thin section analysis and high pressure mercury injection data, the four characteristics of reservoir are analyzed, a more scientific and reasonable calculation model of reservoir logging parameters is established, and the reserves are recalculated after the second interpretation standard of logging is determined. The research improves the accuracy of logging interpretation and provides an effective basis for subsequent production development and potential horizons.展开更多
Accurate estimation of mineralogy from geophysical well logs is crucial for characterizing geological formations,particularly in hydrocarbon exploration,CO_(2) sequestration,and geothermal energy development.Current t...Accurate estimation of mineralogy from geophysical well logs is crucial for characterizing geological formations,particularly in hydrocarbon exploration,CO_(2) sequestration,and geothermal energy development.Current techniques,such as multimineral petrophysical analysis,offer details into mineralogical distribution.However,it is inherently time-intensive and demands substantial geological expertise for accurate model evaluation.Furthermore,traditional machine learning techniques often struggle to predict mineralogy accurately and sometimes produce estimations that violate fundamental physical principles.To address this,we present a new approach using Physics-Integrated Neural Networks(PINNs),that combines data-driven learning with domain-specific physical constraints,embedding petrophysical relationships directly into the neural network architecture.This approach enforces that predictions adhere to physical laws.The methodology is applied to the Broom Creek Deep Saline aquifer,a CO_(2) sequestration site in the Williston Basin,to predict the volumes of key mineral constituents—quartz,dolomite,feldspar,anhydrite,illite—along with porosity.Compared to traditional artificial neural networks (ANN),the PINN approach demonstrates higher accuracy and better generalizability,significantly enhancing predictive performance on unseen well datasets.The average mean error across the three blind wells is 0.123 for ANN and 0.042 for PINN,highlighting the superior accuracy of the PINN approach.This method reduces uncertainties in reservoir characterization by improving the reliability of mineralogy and porosity predictions,providing a more robust tool for decision-making in various subsurface geoscience applications.展开更多
Forests promote the conservation of biodiversity and also play a crucial role in safeguarding theenvironment against erosion,landslides,and climate change.However,illegal logging remains a significant threatworldwide,...Forests promote the conservation of biodiversity and also play a crucial role in safeguarding theenvironment against erosion,landslides,and climate change.However,illegal logging remains a significant threatworldwide,necessitating the development of automatic logging detection systems in forests.This paper proposesthe use of long-range,low-powered,and smart Internet of Things(IoT)nodes to enhance forest monitoringcapabilities.The research framework involves developing IoT devices for forest sound classification andtransmitting each node’s status to a gateway at the forest base station,which further sends the obtained datathrough cellular connectivity to a cloud server.The key issues addressed in this work include sensor and boardselection,Machine Learning(ML)model development for audio classification,TinyML implementation on amicrocontroller,choice of communication protocol,gateway selection,and power consumption optimization.Unlike the existing solutions,the developed node prototype uses an array of two microphone sensors forredundancy,and an ensemble network consisting of Long Short-Term Memory(LSTM)and ConvolutionalNeural Network(CNN)models for improved classification accuracy.The model outperforms LSTM and CNNmodels when used independently and also gave 88%accuracy after quantization.Notably,this solutiondemonstrates cost efficiency and high potential for scalability.展开更多
We propose an integrated method of data-driven and mechanism models for well logging formation evaluation,explicitly focusing on predicting reservoir parameters,such as porosity and water saturation.Accurately interpr...We propose an integrated method of data-driven and mechanism models for well logging formation evaluation,explicitly focusing on predicting reservoir parameters,such as porosity and water saturation.Accurately interpreting these parameters is crucial for effectively exploring and developing oil and gas.However,with the increasing complexity of geological conditions in this industry,there is a growing demand for improved accuracy in reservoir parameter prediction,leading to higher costs associated with manual interpretation.The conventional logging interpretation methods rely on empirical relationships between logging data and reservoir parameters,which suffer from low interpretation efficiency,intense subjectivity,and suitability for ideal conditions.The application of artificial intelligence in the interpretation of logging data provides a new solution to the problems existing in traditional methods.It is expected to improve the accuracy and efficiency of the interpretation.If large and high-quality datasets exist,data-driven models can reveal relationships of arbitrary complexity.Nevertheless,constructing sufficiently large logging datasets with reliable labels remains challenging,making it difficult to apply data-driven models effectively in logging data interpretation.Furthermore,data-driven models often act as“black boxes”without explaining their predictions or ensuring compliance with primary physical constraints.This paper proposes a machine learning method with strong physical constraints by integrating mechanism and data-driven models.Prior knowledge of logging data interpretation is embedded into machine learning regarding network structure,loss function,and optimization algorithm.We employ the Physically Informed Auto-Encoder(PIAE)to predict porosity and water saturation,which can be trained without labeled reservoir parameters using self-supervised learning techniques.This approach effectively achieves automated interpretation and facilitates generalization across diverse datasets.展开更多
Pharmaceutical pollution is becoming an increasing threat to aquatic environments since inactive compounds do not break down,and the drug products are accumulated in living organisms.The ability of a drug to dissolve ...Pharmaceutical pollution is becoming an increasing threat to aquatic environments since inactive compounds do not break down,and the drug products are accumulated in living organisms.The ability of a drug to dissolve in water(i.e.,LogS)is an important parameter for assessing a drug’s environmental fate,biovailability,and toxicity.LogS is typically measured in a laboratory setting,which can be costly and time-consuming,and does not provide the opportunity to conduct large-scale analyses.This research develops and evaluates machine learning models that can produce LogS estimates and may improve the environmental risk assessments of toxic pharmaceutical pollutants.We used a dataset from the ChEMBL database that contained 8832 molecular compounds.Various data preprocessing and cleaning techniques were applied(i.e.,removing the missing values),we then recorded chemical properties by normalizing and,even,using some feature selection techniques.We evaluated logS with a total of several machine learning and deep learning models,including;linear regression,random forests(RF),support vector machines(SVM),gradient boosting(GBM),and artificial neural networks(ANNs).We assessed model performance using a series of metrics,including root mean square error(RMSE)and mean absolute error(MAE),as well as the coefficient of determination(R^(2)).The findings show that the Least Angle Regression(LAR)model performed the best with an R^(2) value close to 1.0000,confirming high predictive accuracy.The OMP model performed well with good accuracy(R^(2)=0.8727)while remaining computationally cheap,while other models(e.g.,neural networks,random forests)performed well but were too computationally expensive.Finally,to assess the robustness of the results,an error analysis indicated that residuals were evenly distributed around zero,confirming the results from the LAR model.The current research illustrates the potential of AI in anticipating drug solubility,providing support for green pharmaceutical design and environmental risk assessment.Future work should extend predictions to include degradation and toxicity to enhance predictive power and applicability.展开更多
This study explores the application of machine learning techniques for predicting permeability,porosity,and flow zone indicator(FZI)in carbonate reservoirs using well log data,aiming to overcome the limitations of tra...This study explores the application of machine learning techniques for predicting permeability,porosity,and flow zone indicator(FZI)in carbonate reservoirs using well log data,aiming to overcome the limitations of traditional empirical methods.Six machine learning algorithms are utilized:support vector machine(SVM),backpropagation(BP)neural network,gaussian process regression(GPR),extreme gradient boosting(XGBoost),K-nearest neighbor(KNN),and random forest(RF).The methodology involves classifying pore-permeability types based on the flow index,leveraging logging curves and geological data.Models are trained using seven logging parameters—spectral gamma rays(SGR),uranium-free gamma rays(CGR),photoelectric absorption cross-section index(PE),lithologic density(RHOB),acoustic transit time(DT),neutron porosity(NPHI),and formation true resistivity(RT)—along with corresponding physical property labels.Machine learning models are trained and evaluated to predict carbonate rock properties.The results demonstrate that GPR achieves the highest accuracy in porosity prediction,with a coefficient of determination(R~2)value of 0.7342,while RF proves to be the most accurate for permeability prediction.Despite these improvements,accurately predicting lowpermeability zones in heterogeneous carbonate rocks remains a significant challenge.Application of cross-validation techniques optimized the performance of GPR,resulting in an accuracy index(ACI)value of 0.9699 for porosity prediction.This study provides a novel framework that leverages machine learning techniques to improve the characterization of carbonate reservoirs.展开更多
As computer data grows exponentially,detecting anomalies within system logs has become increasingly important.Current research on log anomaly detection largely depends on log templates derived from log parsing.Word em...As computer data grows exponentially,detecting anomalies within system logs has become increasingly important.Current research on log anomaly detection largely depends on log templates derived from log parsing.Word embedding is utilized to extract information from these templates.However,this method neglects a portion of the content within the logs and confronts the challenge of data imbalance among various log template types after parsing.Currently,specialized research on data imbalance across log template categories remains scarce.A dual-attention-based log anomaly detection model(LogDA),which leveraged data imbalance,was proposed to address these issues in the work.The LogDA model initially utilized a pre-trained model to extract semantic embedding from log templates.Besides,the similarity between embedding was calculated to discern the relationships among the various templates.Then,a Transformer model with a dual-attention mechanism was constructed to capture positional information and global dependencies.Compared to multiple baseline experiments across three public datasets,the proposed approach could improve precision,recall,and F1 scores.展开更多
基金supported By Grant (PLN2022-14) of State Key Laboratory of Oil and Gas Reservoir Geology and Exploitation (Southwest Petroleum University)。
文摘Well logging technology has accumulated a large amount of historical data through four generations of technological development,which forms the basis of well logging big data and digital assets.However,the value of these data has not been well stored,managed and mined.With the development of cloud computing technology,it provides a rare development opportunity for logging big data private cloud.The traditional petrophysical evaluation and interpretation model has encountered great challenges in the face of new evaluation objects.The solution research of logging big data distributed storage,processing and learning functions integrated in logging big data private cloud has not been carried out yet.To establish a distributed logging big-data private cloud platform centered on a unifi ed learning model,which achieves the distributed storage and processing of logging big data and facilitates the learning of novel knowledge patterns via the unifi ed logging learning model integrating physical simulation and data models in a large-scale functional space,thus resolving the geo-engineering evaluation problem of geothermal fi elds.Based on the research idea of“logging big data cloud platform-unifi ed logging learning model-large function space-knowledge learning&discovery-application”,the theoretical foundation of unified learning model,cloud platform architecture,data storage and learning algorithm,arithmetic power allocation and platform monitoring,platform stability,data security,etc.have been carried on analysis.The designed logging big data cloud platform realizes parallel distributed storage and processing of data and learning algorithms.The feasibility of constructing a well logging big data cloud platform based on a unifi ed learning model of physics and data is analyzed in terms of the structure,ecology,management and security of the cloud platform.The case study shows that the logging big data cloud platform has obvious technical advantages over traditional logging evaluation methods in terms of knowledge discovery method,data software and results sharing,accuracy,speed and complexity.
基金co-funded by the National Key Research and Development Program of China under Grant (2019YFA0708301)the CAS Project for Young Scientists in Basic Research (Grant No.YSBR-082)Research Instrument and Equipment Development Project of Chinese Academy of Sciences (GJJSTD20210008)。
文摘Electromagnetic technology used in logging while drilling(LWD) provides the resistivity distribution around a borehole within a range of several tens of meters.However,a blind zone appears in front of the drill bit when operating in high-angle wells,limiting the ability to detect formations ahead of the drill bit.Look-ahead technology addresses this issue and substantially enhances the proactive capability of geological directional drilling.In this study,we examine the detection capabilities of various component combinations of magnetic dipole antenna.Based on the sensitivity of each component to the axial information,a coaxial component is selected as a boundary indicator.We investigate the impact of various factors,such as frequency and transmitter and receiver(TR) distance,under different geological models.This study proposes 5 and 20 kHz as appropriate frequencies,and 10-14 and 12-17 m as suitable TR distance combinations.The accuracy of the numerical calculation results is verified via air-sea testing,confirming the instrument's detection capability.A test model that eliminated the influence of environmental factors and seawater depth is developed.The results have demonstrated that the tool can recognize the interface between layers up to 21.6 m ahead.It provides a validation idea for the design of new instruments as well as the validation of detection capabilities.
文摘China,as the world’s largest coal producer and consumer,faces increasingly severe challenges from coal mine goaf areas formed through decades of intensive mining.These underground voids,resulting from exhausted resources or technical limitations,not only cause environmental issues like land subsidence and groundwater contamination but also pose critical safety risks for ongoing mining operations,including water inrushes,gas outbursts,and roof collapses.Conventional geophysical methods such as seismic surveys and electromagnetic detection demonstrate limited effectiveness in complex geological conditions due to susceptibility to electrical heterogeneity,electromagnetic interference,and interpretation ambiguities.This study presents an innovative integrated approach combining the Audio-Frequency Electrical Transillumination(AFET)method with multi-parameter borehole logging to establish a three-dimensional detection system.The AFET technique employs 0.1–10 kHz electromagnetic waves to identify electrical anomalies associated with goafs,enabling extensive horizontal scanning.This is complemented by vertical high-resolution profiling through borehole measurements of resistivity,spontaneous potential,and acoustic velocity.Field applications in Shanxi Province’s typical coal mines achieved breakthrough results:Using a grid-drilling pattern(15 m spacing,300 m depth),the method successfully detected three concealed goafs missed by conventional approaches,with spatial positioning errors under 0.5 m.Notably,it accurately identified two un-collapsed water-filled cavities.This surface-borehole synergistic approach overcomes single-method limitations,enhancing goaf detection accuracy to over 92%.The technique provides reliable technical support for safe mining practices and represents significant progress in precise detection of hidden geological hazards in Chinese coal mines,offering valuable insights for global mining geophysics.
基金Supported by the National Natural Science Foundation of China(52288101)National Key R&D Program of China(2024YFF1500600)。
文摘To improve the accuracy and generalization of well logging curve reconstruction,this paper proposes an artificial intelligence large language model“Gaia”and conducts model evaluation experiments.By fine-tuning the pre-trained large language model,the Gaia significantly improved its ability in extracting sequential patterns and spatial features from well-log curves.Leveraging the adapter method for fine-tuning,this model required training only about 1/70 of its original parameters,greatly improving training efficiency.Comparative experiments,ablation experiments,and generalization experiments were designed and conducted using well-log data from 250 wells.In the comparative experiment,the Gaia model was benchmarked against cutting-edge small deep learning models and conventional large language models,demonstrating that the Gaia model reduced the mean absolute error(MAE)by at least 20%.In the ablation experiments,the synergistic effect of the Gaia model's multiple components was validated,with its MAE being at least 30%lower than that of single-component models.In the generalization experiments,the superior performance of the Gaia model in blind-well predictions was further confirmed.Compared to traditional models,the Gaia model is significantly superior in accuracy and generalization for logging curve reconstruction,fully showcasing the potential of large language models in the field of well-logging.This provides a new approach for future intelligent logging data processing.
基金supported by the National Natural Science Foundation of China(No.31660234).
文摘Global forest cover is undergoing significant transformations due to anthropogenic activities and natural disturbances,profoundly impacting hydrological processes.However,the inherent spatial heterogeneity within watersheds leads to varied hydrological responses across spatiotemporal scales,challenging comprehensive assessment of logging impacts at the watershed scale.Here,we developed multiple forest logging scenarios using the soil and water assessment tool(SWAT)model for the Le'an River watershed,a 5,837 km2 subtropical watershed in China,to quantify the hydrological effects of forest logging across different spatiotemporal scales.Our results demonstrate that increasing forest logging ratios from 1.54% to 9.25% consistently enhanced ecohydrological sensitivity.However,sensitivity varied across spatiotemporal scales,with the rainy season(15.30%-15.81%)showing higher sensitivity than annual(11.56%-12.07%)and dry season(3.38%-5.57%)periods.Additionally,the ecohydrological sensitivity of logging varied significantly across the watershed,with midstream areas exhibiting the highest sensitivity(13.13%-13.25%),followed by downstream(11.87%-11.98%)and upstream regions(9.96%-10.05%).Furthermore,the whole watershed exhibited greater hydrological resilience to logging compared to upstream areas,with attenuated runoff changes due to scale effects.Scale effects were more pronounced during dry seasons((-8.13 to -42.13)×10^(4) m^(3)·month^(-1))than in the rainy season((-11.11 to -26.65)×10^(4) m^(3)·month^(-1)).These findings advance understanding of logging impacts on hydrology across different spatiotemporal scales in subtropical regions,providing valuable insights for forest management under increasing anthropogenic activities and climate change.
基金supported by the CNPC Advanced Fundamental Research Projects(No.2023ycq06).
文摘Well logging curves serve as indicators of strata attribute changes and are frequently utilized for stratigraphic analysis and comparison.Deep learning,known for its robust feature extraction capabilities,has seen continuous adoption by scholars in the realm of well logging stratigraphic correlation tasks.Nonetheless,current deep learning algorithms often struggle to accurately capture feature changes occurring at layer boundaries within the curves.Moreover,when faced with data imbalance issues,neural networks encounter challenges in accurately modeling the one-hot encoded curve stratification positions,resulting in significant deviations between predicted and actual stratification positions.Addressing these challenges,this study proposes a novel well logging curve stratigraphic comparison algorithm based on uniformly distributed soft labels.In the training phase,a label smoothing loss function is introduced to comprehensively account for the substantial loss stemming from data imbalance and to consider the similarity between diff erent layer data.Concurrently,spatial attention and channel attention mechanisms are incorporated into the shallow and deep encoder stages of U²-Net,respectively,to better focus on changes in stratification positions.During the prediction phase,an optimized confidence threshold algorithm is proposed to constrain stratification results and solve the problem of reduced prediction accuracy because of occasional layer repetition.The proposed method is applied to real-world well logging data in oil fields.Quantitative evaluation results demonstrate that within error ranges of 1,2,and 3 m,the accuracy of well logging curve stratigraphic division reaches 87.27%,92.68%,and 95.08%,respectively,thus validating the eff ectiveness of the algorithm presented in this paper.
文摘On March 3,2024,the prototype permeability logging instrument independently developed in China successfully completed its first downhole test in Ren 91 standard well in PetroChina Huabei Oilfield.In the open hole section at a depth of 3925 metres and at a temperature of 148℃,the device collected high-quality permeability logging data.This marks a key technological breakthrough from 0 to 1 in permeability logging,and lays the foundation for the next step in developing a complete set of permeability logging equipment.
基金supported by the National Natural Science Foundation of China(No.U21B2062)the Natural Science Foundation of Hubei Province(No.2023AFB307)。
文摘Identification of reservoir types in deep carbonates has always been a great challenge due to complex logging responses caused by the heterogeneous scale and distribution of storage spaces.Traditional cross-plot analysis and empirical formula methods for identifying reservoir types using geophysical logging data have high uncertainty and low efficiency,which cannot accurately reflect the nonlinear relationship between reservoir types and logging data.Recently,the kernel Fisher discriminant analysis(KFD),a kernel-based machine learning technique,attracts attention in many fields because of its strong nonlinear processing ability.However,the overall performance of KFD model may be limited as a single kernel function cannot simultaneously extrapolate and interpolate well,especially for highly complex data cases.To address this issue,in this study,a mixed kernel Fisher discriminant analysis(MKFD)model was established and applied to identify reservoir types of the deep Sinian carbonates in central Sichuan Basin,China.The MKFD model was trained and tested with 453 datasets from 7 coring wells,utilizing GR,CAL,DEN,AC,CNL and RT logs as input variables.The particle swarm optimization(PSO)was adopted for hyper-parameter optimization of MKFD model.To evaluate the model performance,prediction results of MKFD were compared with those of basic-kernel based KFD,RF and SVM models.Subsequently,the built MKFD model was applied in a blind well test,and a variable importance analysis was conducted.The comparison and blind test results demonstrated that MKFD outperformed traditional KFD,RF and SVM in the identification of reservoir types,which provided higher accuracy and stronger generalization.The MKFD can therefore be a reliable method for identifying reservoir types of deep carbonates.
基金Supported by projects of the National Natural Science Foundatio n of China(Nos.41972313,41790453).
文摘Volcanic oil and gas reservoirs are generally buried deep,which leads to a high whole-well coring cost,and the degree of development and size of reservoirs are controlled by volcanic facies.Therefore,accurately identifying volcanic facies by logging curves not only provides the basis of volcanic reservoir prediction but also saves costs during exploration.The Songliao Basin is a‘fault-depression superimposed’composite basin with a typical binary filling structure.Abundant types of volcanic lithologies and facies are present in the Lishu fault depression.Volcanic activity is frequent during the sedimentary period of the Huoshiling Formation.Through systematic petrographic identification of the key exploratory well(SN165C)of the Lishu fault-depression,which is a whole-well core,it is found that the Huoshiling Formation in SN165C contains four facies and six subfacies,including the volcanic conduit facies(crypto explosive breccia subfacies),explosive facies(pyroclastic flow and thermal wave base subfacies),effusive facies(upper and lower subfacies),and volcanogenic sedimentary facies(pyroclastic sedimentary subfacies).Combining core,thin section,and logging data,the authors established identification markers and petrographic chart logging phases,and also interpreted the longitudinal variation in volcanic petro-graphic response characteristics to make the charts more applicable to this area's volcanic petrographic interpretation of the Huoshiling Formation.These charts can provide a basis for the further exploration and development of volcanic oil and gas in this area.
文摘Logging data and its interpretation results are one of the most important basic data for understanding reservoirs and oilfield development. Standardized and unified logging interpretation results play a decisive role in fine reservoir description and reservoir development. Aiming at the problem of the conflict between the development effect and the initial interpretation result of Yan 9 reservoir in Hujianshan area of Ordos Basin, by combining the current well production performance, logging, oil test, production test and other data, on the basis of making full use of core, coring, logging, thin section analysis and high pressure mercury injection data, the four characteristics of reservoir are analyzed, a more scientific and reasonable calculation model of reservoir logging parameters is established, and the reserves are recalculated after the second interpretation standard of logging is determined. The research improves the accuracy of logging interpretation and provides an effective basis for subsequent production development and potential horizons.
基金the North Dakota Industrial Commission (NDIC) for their financial supportprovided by the University of North Dakota Computational Research Center。
文摘Accurate estimation of mineralogy from geophysical well logs is crucial for characterizing geological formations,particularly in hydrocarbon exploration,CO_(2) sequestration,and geothermal energy development.Current techniques,such as multimineral petrophysical analysis,offer details into mineralogical distribution.However,it is inherently time-intensive and demands substantial geological expertise for accurate model evaluation.Furthermore,traditional machine learning techniques often struggle to predict mineralogy accurately and sometimes produce estimations that violate fundamental physical principles.To address this,we present a new approach using Physics-Integrated Neural Networks(PINNs),that combines data-driven learning with domain-specific physical constraints,embedding petrophysical relationships directly into the neural network architecture.This approach enforces that predictions adhere to physical laws.The methodology is applied to the Broom Creek Deep Saline aquifer,a CO_(2) sequestration site in the Williston Basin,to predict the volumes of key mineral constituents—quartz,dolomite,feldspar,anhydrite,illite—along with porosity.Compared to traditional artificial neural networks (ANN),the PINN approach demonstrates higher accuracy and better generalizability,significantly enhancing predictive performance on unseen well datasets.The average mean error across the three blind wells is 0.123 for ANN and 0.042 for PINN,highlighting the superior accuracy of the PINN approach.This method reduces uncertainties in reservoir characterization by improving the reliability of mineralogy and porosity predictions,providing a more robust tool for decision-making in various subsurface geoscience applications.
基金funded by Climate Change AI(2023 innovation grant-https://www.climatechange.ai/innovation_grants).
文摘Forests promote the conservation of biodiversity and also play a crucial role in safeguarding theenvironment against erosion,landslides,and climate change.However,illegal logging remains a significant threatworldwide,necessitating the development of automatic logging detection systems in forests.This paper proposesthe use of long-range,low-powered,and smart Internet of Things(IoT)nodes to enhance forest monitoringcapabilities.The research framework involves developing IoT devices for forest sound classification andtransmitting each node’s status to a gateway at the forest base station,which further sends the obtained datathrough cellular connectivity to a cloud server.The key issues addressed in this work include sensor and boardselection,Machine Learning(ML)model development for audio classification,TinyML implementation on amicrocontroller,choice of communication protocol,gateway selection,and power consumption optimization.Unlike the existing solutions,the developed node prototype uses an array of two microphone sensors forredundancy,and an ensemble network consisting of Long Short-Term Memory(LSTM)and ConvolutionalNeural Network(CNN)models for improved classification accuracy.The model outperforms LSTM and CNNmodels when used independently and also gave 88%accuracy after quantization.Notably,this solutiondemonstrates cost efficiency and high potential for scalability.
基金supported by National Key Research and Development Program (2019YFA0708301)National Natural Science Foundation of China (51974337)+2 种基金the Strategic Cooperation Projects of CNPC and CUPB (ZLZX2020-03)Science and Technology Innovation Fund of CNPC (2021DQ02-0403)Open Fund of Petroleum Exploration and Development Research Institute of CNPC (2022-KFKT-09)
文摘We propose an integrated method of data-driven and mechanism models for well logging formation evaluation,explicitly focusing on predicting reservoir parameters,such as porosity and water saturation.Accurately interpreting these parameters is crucial for effectively exploring and developing oil and gas.However,with the increasing complexity of geological conditions in this industry,there is a growing demand for improved accuracy in reservoir parameter prediction,leading to higher costs associated with manual interpretation.The conventional logging interpretation methods rely on empirical relationships between logging data and reservoir parameters,which suffer from low interpretation efficiency,intense subjectivity,and suitability for ideal conditions.The application of artificial intelligence in the interpretation of logging data provides a new solution to the problems existing in traditional methods.It is expected to improve the accuracy and efficiency of the interpretation.If large and high-quality datasets exist,data-driven models can reveal relationships of arbitrary complexity.Nevertheless,constructing sufficiently large logging datasets with reliable labels remains challenging,making it difficult to apply data-driven models effectively in logging data interpretation.Furthermore,data-driven models often act as“black boxes”without explaining their predictions or ensuring compliance with primary physical constraints.This paper proposes a machine learning method with strong physical constraints by integrating mechanism and data-driven models.Prior knowledge of logging data interpretation is embedded into machine learning regarding network structure,loss function,and optimization algorithm.We employ the Physically Informed Auto-Encoder(PIAE)to predict porosity and water saturation,which can be trained without labeled reservoir parameters using self-supervised learning techniques.This approach effectively achieves automated interpretation and facilitates generalization across diverse datasets.
文摘Pharmaceutical pollution is becoming an increasing threat to aquatic environments since inactive compounds do not break down,and the drug products are accumulated in living organisms.The ability of a drug to dissolve in water(i.e.,LogS)is an important parameter for assessing a drug’s environmental fate,biovailability,and toxicity.LogS is typically measured in a laboratory setting,which can be costly and time-consuming,and does not provide the opportunity to conduct large-scale analyses.This research develops and evaluates machine learning models that can produce LogS estimates and may improve the environmental risk assessments of toxic pharmaceutical pollutants.We used a dataset from the ChEMBL database that contained 8832 molecular compounds.Various data preprocessing and cleaning techniques were applied(i.e.,removing the missing values),we then recorded chemical properties by normalizing and,even,using some feature selection techniques.We evaluated logS with a total of several machine learning and deep learning models,including;linear regression,random forests(RF),support vector machines(SVM),gradient boosting(GBM),and artificial neural networks(ANNs).We assessed model performance using a series of metrics,including root mean square error(RMSE)and mean absolute error(MAE),as well as the coefficient of determination(R^(2)).The findings show that the Least Angle Regression(LAR)model performed the best with an R^(2) value close to 1.0000,confirming high predictive accuracy.The OMP model performed well with good accuracy(R^(2)=0.8727)while remaining computationally cheap,while other models(e.g.,neural networks,random forests)performed well but were too computationally expensive.Finally,to assess the robustness of the results,an error analysis indicated that residuals were evenly distributed around zero,confirming the results from the LAR model.The current research illustrates the potential of AI in anticipating drug solubility,providing support for green pharmaceutical design and environmental risk assessment.Future work should extend predictions to include degradation and toxicity to enhance predictive power and applicability.
基金funded by Fundamental Research Funds for the Central Universities(No.00007851)。
文摘This study explores the application of machine learning techniques for predicting permeability,porosity,and flow zone indicator(FZI)in carbonate reservoirs using well log data,aiming to overcome the limitations of traditional empirical methods.Six machine learning algorithms are utilized:support vector machine(SVM),backpropagation(BP)neural network,gaussian process regression(GPR),extreme gradient boosting(XGBoost),K-nearest neighbor(KNN),and random forest(RF).The methodology involves classifying pore-permeability types based on the flow index,leveraging logging curves and geological data.Models are trained using seven logging parameters—spectral gamma rays(SGR),uranium-free gamma rays(CGR),photoelectric absorption cross-section index(PE),lithologic density(RHOB),acoustic transit time(DT),neutron porosity(NPHI),and formation true resistivity(RT)—along with corresponding physical property labels.Machine learning models are trained and evaluated to predict carbonate rock properties.The results demonstrate that GPR achieves the highest accuracy in porosity prediction,with a coefficient of determination(R~2)value of 0.7342,while RF proves to be the most accurate for permeability prediction.Despite these improvements,accurately predicting lowpermeability zones in heterogeneous carbonate rocks remains a significant challenge.Application of cross-validation techniques optimized the performance of GPR,resulting in an accuracy index(ACI)value of 0.9699 for porosity prediction.This study provides a novel framework that leverages machine learning techniques to improve the characterization of carbonate reservoirs.
基金funded by the Hainan Provincial Natural Science Foundation Project(Grant No.622RC675)the National Natural Science Foundation of China(Grant No.62262019).
文摘As computer data grows exponentially,detecting anomalies within system logs has become increasingly important.Current research on log anomaly detection largely depends on log templates derived from log parsing.Word embedding is utilized to extract information from these templates.However,this method neglects a portion of the content within the logs and confronts the challenge of data imbalance among various log template types after parsing.Currently,specialized research on data imbalance across log template categories remains scarce.A dual-attention-based log anomaly detection model(LogDA),which leveraged data imbalance,was proposed to address these issues in the work.The LogDA model initially utilized a pre-trained model to extract semantic embedding from log templates.Besides,the similarity between embedding was calculated to discern the relationships among the various templates.Then,a Transformer model with a dual-attention mechanism was constructed to capture positional information and global dependencies.Compared to multiple baseline experiments across three public datasets,the proposed approach could improve precision,recall,and F1 scores.
