458 results about "Model building" patented technology

A dynamic optimization system for energy consumption of a refrigeration house based on digital twinning is characterized by comprising a data acquisition module used for acquiring basic structure data of the refrigeration house, technical parameters of a refrigeration system, real-time operation data and historical operation data, preprocessing the data and then outputting a standardized multi-dimensional real-time data stream; the model construction module is used for constructing a 3D geometric model, a thermodynamic transfer model and a refrigeration system mathematical model according to the multi-dimensional real-time data flow, performing machine learning calibration on model parameters through historical operation data, and performing fusion to construct a refrigeration house digital twin model; the prediction analysis module is used for predicting future energy consumption demand and load change according to the refrigeration house digital twin model and the real-time operation data, and outputting an energy consumption prediction result and a load analysis report; a strategy generation module; an execution feedback module; and a learning optimization module. Overall energy consumption of the refrigeration house is reduced, energy utilization efficiency is remarkably improved, and goods storage safety is guaranteed.

The invention provides a construction method of a prediction model based on dynamic gating and cross-modal attention fusion, and belongs to the technical field of model construction. Comprising the following steps: constructing a feature coding layer, and respectively coding multi-source input data to obtain each modal feature vector; constructing a multi-source data interaction layer, and performing deep interaction on each modal feature vector; and finally, carrying out weighted fusion on the main modal features and the cross-modal interaction features based on a dynamic gating mechanism. Constructing a feature fusion layer, and performing time sequence pooling and full-connection fusion on the interacted multi-modal features to obtain a fusion feature vector; and constructing a quantile regression layer, and outputting a prediction result based on the fusion feature vector. According to the method, the prediction model is constructed by fusing the dynamic gating mechanism and the cross-modal attention mechanism, so that the problems of an existing prediction model in the aspects of deep fusion of multi-source heterogeneous data, cross-modal dynamic interaction and accurate quantification of tail risks are solved, and then efficient prediction of stock price collapse risks is realized.

The invention provides a digital twin three-dimensional model construction method based on a generative large model, and the method comprises the steps: obtaining multi-source monitoring data of a power distribution network, and processing the multi-source monitoring data into a training data set; the method comprises the following steps: mapping multi-source monitoring data into a multi-scale tensor subspace through tensor wavelet structured transformation, adaptively extracting spatial features through a learnable wavelet kernel, and keeping the structural continuity of a physical field in combination with a geometric prior regular term; constructing and training a generative adversarial network through a training data set; inputting and analyzing the physical parameter vector of the target scene, and if the topological similarity score is lower than a preset threshold value, adjusting noise vector regeneration; and if yes, outputting a three-dimensional model tensor and importing the three-dimensional model tensor into a digital twin platform, and driving real-time physical field visualization. According to the method, characteristics of a multi-scale space structure and a nonlinear physical field can be reserved, the physical rationality and generalization ability of the generated model are remarkably improved, and depth identification of topological attributes (such as hole connectivity and surface defects) and local geometric defects of the three-dimensional model is realized.

The invention provides a construction method and system of a rock-soil shear strength parameter prediction model, and relates to the technical field of model construction.The construction method comprises the steps that CT scanning data of a real rock-soil sample is obtained and preprocessed, and an irregular three-dimensional particle set for geometric modeling is obtained; performing particle surface roughness modeling analysis on the three-dimensional particle body set to obtain a surface roughness coefficient set of the three-dimensional particle body set; performing particle size cumulative distribution function segmentation processing and fractal dimension control on the surface roughness coefficient set to obtain a structure grading sample set; constructing a THMC multi-field coupling model based on the structure grading sample set to obtain a working condition particulate matter physical response parameter set; and performing a simulation experiment on the working condition particle physical response parameter set, and performing deep neural network modeling processing based on an experiment result to obtain a strength response model capable of predicting the cohesive force and the internal friction angle of the rock-soil body under any working condition input. According to the invention, the prediction efficiency and accuracy are significantly improved.

The invention provides a deep tunnel excavation physical simulation test three-dimensional stress field real-time prediction method, and relates to the technical field of tunnel engineering. Comprising two stages of model construction and training and model use. In the model construction and training stage, a physical constraint reference model is constructed, and training optimization is completed; and in the model use stage, real-time prediction of a stress field of a deep engineering physical simulation test is realized through the trained and optimized model. The method has breakthrough advantages in the aspects of physical law constraint, real-time monitoring data fusion, training loss stability, boundary stress condition accuracy and the like, and is a high-reliability three-dimensional stress field prediction method oriented to a deep complex environment.

The invention discloses a multi-device cooperative control system and method for a construction site intelligent capsule bin, and relates to the technical field of industrial control, and the system comprises a data interface layer, a model construction module, a situation prediction module, a cooperative control unit, a self-adaptive cooperative optimization module, and a safety helmet positioning module. The method comprises a model construction step, a data acquisition step, a simulation prediction step, a situation judgment step, a cooperative control step and a self-adaptive cooperative optimization step. According to the method, unified data fusion and cooperative control are carried out on various devices, dynamic adjustment and strategy optimization of the devices can be carried out in time according to the actual situation of the site, the resource scheduling efficiency is improved, and the utilization efficiency of data is improved.

The invention provides a three-dimensional visualization analysis method for a real scene of an ancient earthquake exploring groove, which belongs to the technical field of seismic geological survey and comprises a data acquisition module used for acquiring point cloud data of target exploring grooves in different regions. And the model building module is used for carrying out scene format conversion on the point cloud data of the target exploring grooves in different regions, and carrying out exploring groove three-dimensional model building by utilizing the converted SLPK file. And the visualization module is used for carrying out live-action three-dimensional display on the three-dimensional models of the target probe grooves in different regions through a three-dimensional scene Web server based on the SLPK files in different regions, and rendering node data of different detail levels of the probe grooves according to user requirements. The ancient earthquake event can be better understood, and a reliable three-dimensional digital file and a new shared analysis normal form are provided for the ancient earthquake.

The invention relates to the technical field of data processing, in particular to a three-dimensional live-action model construction system for building structure demolition and reconstruction. The system extracts event combinations with obvious transfer correlation features for localization and analysis of dangerous sound sources. Through simulation of transmission of test waves in an initial acoustic slowness field, a most probable dangerous sound source position for generating a current event combination under each test wave is obtained by using an objective function optimization solution method. And further performing statistics on simulation information of all test waves, establishing a residual curve, screening out a brittleness event position based on feature extraction, and adjusting a slowness value in a current voxel through various parameters corresponding to the brittleness event position in an initial acoustic slowness field. Through the dynamic correction method, the obtained corrected acoustic slowness field can more effectively represent the specific change of the building structure in the dismantling process, so that the internal damage can be more clearly quantified.

The invention discloses a cross-modal large model construction method and system based on track spatio-temporal characteristics, and belongs to the crossing field of artificial intelligence and dynamic spatio-temporal data processing, and the method comprises the steps: carrying out the sliding sampling and spatial distribution difference judgment through the semantic dynamic segmentation of multi-scale track spatio-temporal data, and generating spatio-temporal data blocks with consistent semantics; designing a space-time encoder of a hybrid architecture, extracting track time sequence association and spatial features, and unifying dimensions; constructing a text space-time fusion mechanism, and dynamically adapting cross-modal features by means of an anchor interface and gating fusion; a staged instruction fine tuning strategy is adopted, semantic alignment of space-time and text features is optimized firstly, then model top-layer parameters are trained cooperatively, complex scene adaptation is enhanced in combination with instruction difficulty progression and hard sample mining, and space-time constraint regular terms are introduced to guarantee output rationality. According to the method, high-precision cross-modal reasoning capability is provided for scenes such as track analysis and track prediction.

The invention relates to the technical field of water resource management and forecasting, in particular to a deep learning model building and forecasting method based on hydrological mechanism fusion. The method specifically comprises the following steps: constructing a snow melting calculation module and a soil calculation module, splicing to form a runoff production model, constructing a confluence calculation module, executing confluence evolution simulation of a calculation result of the runoff production module by using the confluence calculation module, and constructing a confluence model. In confluence model simulation calculation, according to the number of calculation units, a broadcast calculation strategy and an array-based ordinary differential equation solving method are adopted, parallel calculation of the multiple calculation units is achieved, and runoff production calculation results of the multiple units are obtained; and integrating and calculating runoff production calculation results of each unit through a confluence model to obtain a runoff prediction result of the modeled drainage basin after confluence. The method provided by the invention solves the problems of low model construction efficiency and poor practical application effect faced by the application of the deep learning technology in the hydrological model at present.

The invention relates to the technical field of model construction, can be applied to business scenes such as financial science and technology, and discloses a double-track evaluation driven optimization processing method, device and equipment and a medium, and the method comprises the steps: constructing a double-track evaluation system, analyzing training data, generating optimized training data, and building a target model; generating a capability index by utilizing a double-track evaluation system and multi-role agent simulation, and generating an alignment analysis result in combination with reinforcement learning performance; obtaining real service feedback indexes to form a training evaluation index set, and determining an attribution link; and generating training adjustment information based on the attribution link and driving the target model to be retrained to obtain an optimized target model for processing the service input data and outputting a service processing result. According to the method, the evaluation result and the service feedback are linked to realize a training closed loop, so that the capability gap of the model can be identified and corrected, and the compliance, the reliability and the application effect of the model in the actual service are improved.

The invention relates to the technical field of elevator safety monitoring, and discloses an elevator safety detection system based on vision and multi-modal data fusion, comprising: a data preprocessing module for collecting and preprocessing original multi-modal sensor data of elevator operation; the feature extraction module is used for identifying operation conditions and extracting multi-scale features for different conditions; the multi-modal fusion module is used for calculating the reliability coefficient of each modal and carrying out adaptive fusion; the anomaly detection module is used for extracting transient characteristics of working condition switching, carrying out anomaly judgment and mining an anomaly evolution trend; the sample enhancement module is used for extracting normal samples and rare fault samples, generating synthetic samples and performing quality screening; the model construction module is used for training a multi-level fault identifier and fusing a multi-modal feature library to carry out optimization learning; the intelligent early warning module is used for calculating a safety level and reasoning a fault reason and a disposal measure; accurate dynamic safety monitoring of the super high-rise building elevator under the extreme load switching working condition is achieved.

The invention relates to an end-to-end point cloud denoising model construction method and system, belongs to the technical field of deep learning, and solves the problems that an existing denoising model is insufficient in generalization ability and output point cloud geometric features are not high in fidelity. Comprising the steps that after collected clean point clouds are preprocessed, point cloud samples are generated by injecting noise, and a sample set is constructed; constructing a deep learning model, and constructing a combined loss function by fusing multiple supervision signals; the deep learning model sequentially comprises a feature extraction module, a denoising guide prediction module and a point cloud attribute prediction module, and is used for outputting a prediction coordinate and a prediction normal vector of a denoised point cloud; and performing end-to-end training on the deep learning model based on the sample set and the combined loss function, and taking the trained deep learning model as a point cloud denoising model. The generalization ability of the denoising model and the denoising effect are improved.

The invention discloses a material level intelligent sensing and early warning system in the silo unloading process, and relates to the technical field of silo material level intelligent sensing. The method comprises the following steps: a data acquisition module used for acquiring a multi-modal data stream; the model construction module is used for constructing a normal unloading model based on the multi-modal data flow and a preset physical mechanism constraint; the state judgment module is used for extracting a current feature vector from the multi-modal data stream; comparing the current feature vector with a feature distribution range in a multi-dimensional feature space determined by a normal unloading model, and judging a current material flow state; according to the invention, material flow abnormal symptoms can be captured in time, early warning of the blocking risk is realized, the production rhythm is prevented from being disturbed by material flow interruption, the potential safety hazard caused by manual entering into the silo for troubleshooting is reduced, the grain loss and equipment damage risks caused by long-term retention and mildew of materials are reduced, and the storage safety and the operation stability of a grain storage system are guaranteed.

The invention discloses a water quality probability forecasting method based on Bayesian multi-time-sequence deep learning. The method comprises the following steps: S1, determining a forecasted water environment water ecological index, a driving index and a forecasting day number; s2, collecting time sequence data monitored by the forecasting indexes and the driving indexes, and after data preprocessing, constructing a data set required by model construction; s3, carrying out data division on the time sequence data, constructing a driving index forecasting model by adopting a multi-time sequence deep learning method, and carrying out parameter learning by selecting a Bayesian random discarding method; s4, performing effect evaluation on the accuracy and precision of the model, and adopting a hyper-parameter optimization method to improve the simulation forecast effect; s5, carrying out model training by adopting all data without segmenting the training set and the test set, carrying out water quality probability forecasting by utilizing the trained model, and outputting a forecasting mean value and a confidence interval; according to the method, the confidence interval is output while high-precision prediction is provided, and the scientificity and stability of prediction are improved.

The invention relates to the technical field of structural design and optimization, and particularly discloses a structural error tolerance model construction method for compartment connection positioning, which comprises the following steps of: obtaining real point cloud data of a bolt connection pair through three-dimensional laser scanning, and realizing parameterized digital geometric modeling in combination with a deep neural network; material attributes are given, and self-adaptive grid division is carried out; generating a dynamic load sequence based on a vehicle driving condition, applying the dynamic load sequence to the model, and extracting full-field displacement vector data by using a digital image correlation technology; further identifying a micro-slip area of the contact interface, and calculating an energy dissipation index of the micro-slip area; a quantitative mapping relation between energy dissipation and fatigue life is established through experimental calibration, key design parameters are iteratively optimized by adopting a pattern search method with the target life as a constraint, and finally a connection structure scheme meeting the durability requirement is output.

The invention discloses a water conservancy and hydropower equipment simulation test method which is characterized by comprising the following steps: S1, twin construction and data acquisition, S2, twin dynamic evolution, S3, model construction and state prediction, S4, multi-field coupling simulation test, and S5, result analysis and strategy feedback. A water conservancy and hydropower equipment simulation test system is characterized by comprising a digital twinborn body module, an artificial intelligence auxiliary simulation module, a multi-physics field coupling modeling module and a result analysis and feedback module. According to the method, through high-precision modeling and multi-source data fusion, in combination with adaptive filtering and a deep reinforcement learning algorithm, the error between a digital twinborn body and physical equipment is smaller than 2%, the dynamic evolution period is shortened to 3 minutes, the running state of the equipment can be reproduced in real time, a precise data base is provided for fault early warning, the on-site inspection frequency is reduced, and the labor cost is reduced by 30% or above.

The invention provides an existing building BEM modeling method, system and device based on an infrared thermal image of an unmanned aerial vehicle and a medium, and belongs to the technical field of building modeling. Generating a three-dimensional geometric model of the building main body; screening out an external window candidate area by analyzing RGB value characteristics of the visible light point cloud, generating a window opening three-dimensional boundary, and determining an internal surface temperature and an indoor air temperature through internal surface infrared temperature measurement and area weighted integration; and based on the temperature data and the building envelope construction parameters, inverting heat transfer coefficients of the outer wall, the roof and the outer window, and finally constructing a building energy consumption model in combination with the weighted average K value. According to the method, precise operation from building geometric model construction to thermal parameter acquisition and then building energy consumption model construction is realized. The efficiency and precision of BEM modeling of the existing building are improved, and a scientific basis is provided for building energy-saving transformation, energy consumption evaluation and the like.

The invention discloses a geophysical three-dimensional visualization and geologic model construction system, which belongs to the technical field of geophysical data processing and three-dimensional geologic modeling and comprises a multi-source data acquisition module, a volume rendering module, an interactive segmentation module, an uncertainty expression module, a multi-scale fusion module, an attribute interpolation module and a model construction module. Multi-source geophysical data are converted into a visual three-dimensional visual model through a volume rendering technology, an interactive segmentation algorithm is adopted to accurately extract a geologic body boundary, an uncertainty expression mechanism is innovatively introduced to quantify the reliability of the model, and a multi-scale fusion technology is integrated to realize seamless connection from a regional structure to an ore deposit fine structure. The Kriging interpolation functional product physical property parameter three-dimensional distribution field is provided, the spatial cognitive ability and the decision-making efficiency under the complex geological condition are remarkably improved, and a powerful visual analysis platform is provided for geological exploration and resource evaluation.

The invention relates to the technical field of geological engineering, in particular to a tunnel karst development degree intelligent evaluation method and device and a medium, and the method comprises the steps: obtaining a multi-source fusion data set; generating a karst development fractal evaluation comprehensive index; determining a tunnel karst key detection area; constructing a three-dimensional medium distribution model of the karst area; carrying out karst area risk grade classification, and establishing a geological risk area prediction model; according to the method, the comprehensiveness and the accuracy of karst development characteristics are remarkably improved by combining multi-source data and a fractal theory; a detection path is dynamically optimized by using a reinforcement learning model, detection blind areas are reduced, and the detection efficiency of key areas is improved; noise reduction and fusion are performed on the data through the generative adversarial network, so that the accuracy of model construction is enhanced; based on the three-dimensional medium distribution model and the risk prediction model, graded evaluation and dynamic prediction of the risk of the karst area are effectively realized.

The invention discloses a building equipment and material intelligent transportation analysis system based on mixed reality, and relates to the technical field of transportation in the construction process, and the system comprises a data collection module, a virtual model construction module, a transportation path planning module, an equipment disassembly degree measurement and calculation module, a scheme economy evaluation module and an optimal scheme output module. According to the invention, on the basis of mixed reality equipment and an unmanned aerial vehicle telemetering technology, high-efficiency acquisition and high-precision three-dimensional virtual modeling of transportation channel and equipment data are completed, manual measurement limitation is broken through, traffic is dynamically simulated, paths are screened, calculation and disassembly are carried out, the structure safety is guaranteed, and a complex scene is adapted; a cost-benefit balance algorithm is introduced, a multi-dimensional evaluation system is constructed, the cost and influence are comprehensively considered, additional evaluation is dynamically triggered, a special scene is adjustable, and scientific support is provided for intelligent transportation management.

The invention relates to the technical field of data processing, and discloses a three-dimensional drilling data integration and display system for safety evaluation of an earthquake area. Comprising a mode selection module for selecting a point location marking mode of an earthquake region, a point location marking module for marking evaluation data point locations distributed at intervals, a data acquisition module for acquiring three-dimensional drilling data, a model construction module for constructing a region simulation model matched with the earthquake region, and an interactive display module. The dividing module is used for dividing a regional simulation model into sub-models and establishing regional channels between adjacent sub-regions; according to the invention, dynamic integration and display effects of three-dimensional drilling data at adjacent positions of an earthquake area can be realized, and an independent and reliable interactive display channel can be provided for integration and display of three-dimensional drilling data at different time lines and different geographic positions; the targeted integration and display effect of the local position of the earthquake area on the timeline is realized, and the accuracy of safety evaluation of the earthquake area is improved.

The invention discloses an axial flow compressor blade anti-erosion optimization method facing a sand and dust environment, and belongs to the technical field of compressor blade anti-erosion design and optimization. According to the method, on the basis of a gas-solid two-phase flow erosion mechanism in a sand and dust environment, links such as working condition analysis, parameterization platform building, constrained multi-objective optimization method design, model building and optimization and index verification are integrated, and an integrated anti-erosion optimization process is formed. CFturbo, AutoGrid, CFX and Matlab are called in a unified mode through a Python script, automatic coupling of geometric modeling, grid division, gas-solid coupling simulation and an intelligent optimization algorithm of the blade is achieved, and collaborative optimization of aerodynamic performance and erosion resistance can be achieved under the constraint of erosion wear. Compared with a traditional method which depends on experience modification or single simulation, the method has the advantages that the optimization efficiency is remarkably improved while the simulation precision is guaranteed, the personal error risk is reduced, and the method can flexibly adapt to different compressor models and diversified operation conditions.

The invention relates to a differential protection closed-loop simulation method based on a high-performance full-electromagnetic transient simulation platform. The method comprises the steps of model construction, data interaction, closed-loop simulation and performance evaluation. According to the method, modeling is carried out based on a Cloud PSS platform, and full electromagnetic transient elements of a primary model and a secondary model can be considered. Under the background that differential protection of a novel power system and a new-generation ship power supply system faces an adaptability problem, the method can comprehensively and efficiently establish a differential protection closed-loop simulation model, and detailed description of differential protection action performance is achieved; according to the scheme, the influence of primary model modeling granularity on differential protection is fully considered, and accurate description and evaluation of the differential protection adaptability problem under the background of a novel power system and a novel ship power supply system can be realized.

The invention provides a small sample AI proxy model construction method based on sensitivity analysis and supplementary sampling, and the method comprises the steps: carrying out the global sensitivity analysis through the automatic simulation process of parameterized parts of wind power equipment in combination with test design, and removing insensitive variables, thereby guaranteeing the accuracy and standardization of a sample, avoiding the invalid consumption of irrelevant variables, and achieving the automatic simulation of the parameterized parts of the wind power equipment. Latin hypercube sampling is adopted based on key variables, a core design area is uniformly covered, the number of initial simulation times is greatly reduced, the cost is controlled, the problem that the generalization ability is poor due to uneven small sample distribution is solved, a deep neural network regression model is trained through a training set, the strong nonlinear fitting ability of the deep neural network regression model adapts to a complex mapping relation, the core law is efficiently learned, and the robustness is high. Based on initial model prediction precision and sensitivity information, samples are accurately supplemented in weak areas, samples are not increased blindly, investment is reduced, prediction blind areas are made up, errors are gradually reduced through iterative training closed-loop optimization, and finally the small sample AI proxy model meeting preset requirements is obtained.

The invention discloses an automatic shell-making wire environment simulation system. The system comprises an environmental parameter testing unit, a heat and mass transfer model establishing unit and an optimization processing unit. The environmental parameter testing unit is used for testing environmental parameters influencing the drying process of the shell-making drying cabin through testing equipment, evaluating the uniformity of airflow in the drying cabin and the change condition of the environmental parameters, and sending testing and evaluating results to the heat and mass transfer model establishing unit; the environmental parameters at least comprise wind speed. And the heat and mass transfer model building unit is used for building a shell drying heat transfer model and a shell drying mass transfer model, evaluating the shell drying heat transfer model and the shell drying mass transfer model based on the model expression indexes after building is completed, and analyzing the influence of the environmental parameters on shell heat transfer and mass transfer based on the information transmitted by the environmental parameter testing unit. And the optimization processing unit is used for introducing the influence of load, airflow in the drying cabin and air into the established shell drying heat transfer model and shell drying mass transfer model for simulation and optimization.

The invention relates to the technical field of switch cabinet simulation, in particular to a switch cabinet mechanism model construction method, system and device and a medium, and the method comprises the steps: generating a three-dimensional model corresponding to a switch cabinet, carrying out the preprocessing of the three-dimensional model to obtain a simulation model, applying a target parameter in an actual use scene to the simulation model to complete the parameter loading, and obtaining a parameterized model, determining a target relational expression of the switch cabinet, unifying an expression structure and modeling requirements of the switch cabinet, combining the target relational expression with the parameterized model, and generating a mechanism model through modeling processing; simulation adaptability and calculation efficiency are improved through structure simplification and grid division, actual operation parameters are loaded to reflect different working condition characteristics, an electromagnetic field, temperature field and electromagnetic loss multi-physical relational expression is constructed to achieve unified modeling, a mechanism model capable of being used for multiple operation conditions is formed, and the method is suitable for large-scale popularization and application. The method is suitable for application scenes such as switch cabinet design verification, operation maintenance and fault pre-judgment.

The invention discloses a construction safety monitoring system for house construction, which belongs to the technical field of safety monitoring and comprises a construction monitoring data integration module, a construction monitoring image denoising module, a construction safety monitoring model building module and a real-time construction safety monitoring module. The method specifically comprises the following steps: respectively extracting three types of noise features and obtaining a weight coefficient, cascading a double-branch enhanced feature map and connecting the feature map with an initial feature map residual error, fusing a security mask, constructing a three-scale encoder and an asymmetric decoder, mapping a deep fusion feature into a noise feature and removing the noise feature, and providing a clear and complete image data support; dividing personnel feature map windows based on double scales, performing weighted fusion on same-scale features and aligning resolutions, extracting local, mesoscale and global features in parallel, weighting query vectors element by element by using importance weights of all positions, screening effective spatial features through convolution and gating weights, and obtaining a query result; and the real-time performance and the accuracy of house construction safety monitoring are improved.

The invention relates to the technical field of underground engineering construction safety monitoring, and discloses an underground pipe-jacking tunnel excavation pavement settlement prediction system and a prediction method thereof.The prediction system comprises an acquisition terminal and a computing server and is provided with a data preprocessing module, a model building module and other modules. The core lies in that a CA-CNN-LSTM hybrid neural network is constructed, geological spatial features are extracted by using parallel convolutional neural network branches, construction time sequence features are extracted by using bidirectional long short-term memory network branches, and deep fusion of spatial and temporal features is realized through a Cross-Attention module and a dynamic gating unit. The prediction method relates to multi-source data cleaning enhancement, model training and double-index grading early warning. According to the method, the problem of heterogeneous feature coupling is effectively solved, the settlement prediction precision is effectively improved, hierarchical control signals can be generated according to the prediction result, and the construction safety is effectively guaranteed.

The invention discloses a classification method and system and a model construction method. The method comprises the steps of obtaining multi-modal data of a target object in a distributed system, the multi-modal data being used for reflecting a traffic behavior of the target object; a plurality of single-mode feature extractors are adopted to perform feature extraction on the corresponding modes in the multi-mode data to obtain a plurality of single-mode feature vectors, and the single-mode feature extractors perform parameter updating based on a global model issued by a server in the distributed system to obtain a plurality of single-mode feature vectors; the global model is determined based on model parameters of single-mode feature extractors corresponding to a plurality of edge devices in the distributed system; and determining the type of the target object according to the plurality of single-mode feature vectors. The technical problems that user behavior recognition is delayed and privacy disclosure risks exist due to the fact that a central server is adopted to process large-scale user original data in a centralized mode in the related technology are solved.