277 results about "Model building" patented technology

The invention, which belongs to the technical field of model construction, discloses a batch type refined three-dimensional modeling method of a building frame model. The method comprises the following steps: determining a range frame of a batch type refined modeling region of a building; constructing a three-dimensional model; establishing a frame model index; and constructing a model of a batch type refined modeling accessory structure. According to the method, a BSP building model reconstruction method under a DLG building outer contour constraint is put forward; and with the high-plane precision of a DLG, a defect of low precision of a point cloud boundary is overcome, so that strict matching of the main contour of the frame model with the outer contour of the DLG can be realized. On the basis of a one-to-one correspondence relationship among a building DLG base map, a building point type in a range, and a generated model, batch type refined modeling and quality checking of buildings in a large area can be realized and a well-developed production and quality checking working flow can be formed. A model space matching and duplicating method is put forward and base elevation and precise orientation information of the model can be obtained automatically, so that the precise modeling efficiency can be improved.

The invention relates to a novel distributed Hebei model construction method and application thereof. The novel distributed Hebei model construction method comprises the following steps that firstly, drainage basin generalization is carried out; secondly, grids serve as study units, a Hebei model is used for establishing a grid runoff producing module; thirdly, a lag-and-route method and a Muskingum method are used for establishing grid-by-grid distributed confluence modules. The mature Hebei model is used for conducting grid runoff producing calculation, then, grid-by-grid confluence calculation is carried out, the situation that the space distribution of factors such as an underlying surface is uneven is fully considered in the model, the model can be coupled with an atmospheric mode, the forecast accuracy of storm flood is improved, the meeting period is prolonged, and the novel distributed Hebei model construction method has important significance for storm flood forecasting, a drainage basin water circulation mechanism, water resource planning, unified water resource management scheduling in the northern semiarid and semihumid regions in China.

The invention provides a function gradient material hidden model building method based on a distance field. The method comprises the following concrete steps that (1) a geometric distance field is created: an entity prototype is measured, surface point cloud data is obtained, the point cloud data is a position coordinate value, a surrounding box is used for surrounding the position coordinate, the surrounding box is uniformly divided, and signed distance fields of the point cloud data of all dividing lattice top points in the surrounding box are calculated for displaying a geometric model of the entity; (2) a material distance field is created: the material features are determined according to the material design intention of the function gradient materials, so that the material distance field is calculated; (3) function gradient materials are subjected to hidden modeling. The functional gradient material geometric distance field and the material field modeling are creatively combined, and the hidden function mode of the distance field is used for completely expressing the geometric and material information. Through the self-adaptive characteristic of the distance field, the method is applicable to any complicated models, and the Boolean operation can be conveniently carried out.

The invention discloses a method and a system for efficiently estimating the near-surface PM2.5 (particular matter 2.5) concentration. The method includes a model building step and a model estimating step. The model building step further includes a data compressing sub-step, extracting main spectral signal structure characteristics of remotely sensed data; a data matching sub-step, extracting corresponding remotely sensed information according to spatial coordinates of PM2.5 ground monitoring data; a model building sub-step, building estimation models according matched data sets. The model estimating step further includes an estimation requesting sub-step, preprocessing estimation input data; an estimating sub-step, estimating the near-surface PM2.5 concentration according to estimation requests and outputting estimation results. The method and the system have the advantages that the near-ground PM2.5 concentration can be estimated according to the MODIS (moderate resolution imaging spectroradiometer) observation data on the basis of the artificial neural network models, and accordingly remote sensing operational monitoring requirements can be met; the method and the system can support importing of meteorological factors, and accordingly the near-surface PM2.5 concentration can be quickly and efficiently dynamically monitored on a large scale.

The invention discloses a prediction model construction method and device, a prediction method and device, equipment and a medium. The construction method comprises the steps of obtaining a sequence value of a historical power load of a prediction target; based on a preset value range of model parameters, determining the model parameters by utilizing the correlation characteristics of the sequencevalues; and constructing a prediction model based on the model parameters, with the prediction model being used for predicting the power load of the prediction target in the next time period. According to the embodiment of the invention, the obtained historical power load sequence value of the enterprise user is analyzed. , The model parameters are determined according to the model parameters, and then the prediction model is constructed by utilizing the determined model parameters, so that the electricity selling enterprise can accurately predict the medium and long-term power load of the enterprise in the future by utilizing the constructed prediction model, a reliable basis is provided for the electricity selling enterprise in an electricity selling quantity declaration service, and deviation assessment is reduced.

The invention discloses a machine learning platform and a data model optimization method based on the platform, and relates to the technical field of data model evaluation. The system comprises a model construction module, a data acquisition module, a data processing module, a model evaluation module and a model optimization module. The model construction module comprises a model construction unitand a model release unit; The model building unit is used for building a model; And the model publishing unit publishes the constructed data model. According to the invention, the data model is rapidly constructed through the model construction module; The data acquisition module is adopted to acquire the source data, the source data is processed by the data processing module and then is lifted to the data model, the effectiveness of the data entering the data model is guaranteed, model evaluation errors are reduced, and the model evaluation module evaluates the data model by outputting the confusion matrix of the model and the model accuracy, so that the optimization of the data model is facilitated.

The invention provides a three-dimensional model construction method, device and equipment and a computer readable storage medium, and belongs to the technical field of three-dimensional modeling. The method comprises the steps that vector data of each object in a to-be-modeled area is obtained, and the vector data comprises attribute information and space information; based on the spatial information of each object in the to-be-modeled area, geometric structure body sub-models are generated, and one object corresponds to one geometric structure body sub-model; texture mapping processing is carried out on the geometric structure body sub-model of the target object by adopting mapping textures associated with attribute information of the target object, so that a three-dimensional model is obtained, and the to-be-modeled area comprises the target object; and the three-dimensional model is output. According to the invention, the modeling efficiency of three-dimensional modeling can be improved.

The invention discloses a well-to-seismic joint initial lithologic model construction method based on deep learning, is applied to the field of three-dimensional geological modeling, and aims to solvethe problems that in the prior art, the frequency of well logging data is too low due to filtering, a lot of high-frequency effective information is lost, and seismic data cannot be effectively controlled in the interpolation process. According to the invention, the convolutional neural network is used to extract characteristics of long and short periods, namely high and low frequencies, contained in data; different features are classified and learned by adopting a long-term and short-term memory network, so that the relationship between seismic data and logging data is learned accurately, accurate prediction of rock attributes is achieved, a lithology initial model is constructed, a basis is provided for inversion of lithology parameters, and exploration and development of oil and gas and reservoir description of oil and gas reservoirs are guided.

The invention discloses a medium and long term runoff ensemble forecasting method based on multi-model combination, and relates to the technical field of hydrological prediction. According to the method, multiple machine learning algorithms are adopted to construct a medium-and-long-term runoff forecasting model, the medium-and-long-term runoff forecasting model is used as a weak learner, and an integrated model construction method based on multi-model combination is provided on the basis. Meanwhile, equivalent forecast is searched through parameter disturbance to construct a forecast set, andensemble forecast is carried out. Compared with an existing common deterministic forecasting method, the method has the advantages that part of defects existing in the method are improved, and the precision and generalization capacity of medium-and-long-term forecasting are improved. Meanwhile, the uncertainty of forecasting is quantitatively described through probability forecasting, and the accuracy and reference value of forecasting are improved.

The invention discloses a medium-and-long-term rainfall forecast modeling method based on whole-process coupling machine learning, and the method comprises the following steps: S1, data processing: collecting actually measured rainfall, 130 meteorological-climate indexes and other data, and determining a forecast structure; s2, factor screening: providing a factor screening method based on Laplacian fractional-recursive feature elimination, and obtaining a forecast factor set; s3, model construction: constructing a plurality of machine learning models, and solving a plurality of sets of sub-forecasting results by adopting the forecasting structure and the forecasting factor set; and S4, multi-model fusion: providing a multi-model fusion technology based on an improved stacking method, and outputting a final forecast result. According to the method, the latest research result of the machine learning theory is applied to each link of medium and long term rainfall forecasting, the theoretical basis is sufficient, the practical application is reasonable, and the accuracy and the reliability of month-season-year scale rainfall forecasting can be effectively improved.

The invention discloses a drilling data model construction method based on BIM, and the method comprises the steps: building a spatial information database according to various types of exploration data, and then carrying out the building of a drilling data model: reading the exploration data of modeling region surface point cloud data or contour lines from the spatial information database, and automatically building a surface model; reading the drilling data of the modeling area from the spatial information database, and automatically creating a drilling data model; judging whether the groundsurface model conflicts with or is unreasonable with the ground surface generated by the drilling data model or not, and if so, correcting the drilling data of the modeling area; and if not, automatically numbering and loading user-defined attribute information, and outputting / storing the drilling data model. According to the method, automatic modeling is achieved, seamless butt joint of the building BIM model can be achieved, the simulation performance of traditional BIM application such as early-stage planning design, earthwork calculation and site construction simulation can be greatly improved, and the accuracy of geologic body numerical analysis in the engineering geology field is greatly improved.

The invention discloses a boosting algorithm-based construction method and construction system of a Kawasaki disease risk assessment model. The construction method includes the following steps that: valid samples that can be used for modeling assessment are extracted from a sample data set; 10 features that meet on-site medical auxiliary diagnosis application are selected from the feature set of the valid sample; the incomplete data set of the valid samples is randomly divided into a training set and a verification set; a boosting method is adopted to fit the training set, so as to perform model construction, and a ten-fold cross-validation method is adopted to record optimal model parameters; and the verification set is adopted to calculate a model classification threshold t according toan ROC curve, and therefore, the Kawasaki disease risk assessment model is constructed. The invention also provides a corresponding Kawasaki disease risk assessment system for assessing data to be assessed so as to obtain a KDx score. With the system and method of the invention adopted, the rate of misdiagnosis and the rate of missed diagnosis of the Kawasaki disease can be decreased, and patientscan obtain effective prevention, intervention and treatment in the early stage of the disease.

The invention discloses a mass-customization two-dimensional spatiotemporal model modeling and an optimization method. Firstly, the mass-customization two-dimensional spatiotemporal model modeling adopts the time dimension of process model description and the spatial dimension of product model description; the total production time for the order form finished through the mass-customization is mapped to the time dimension; the total production cost for the order form finished through the mass-customization is mapped to the spatial dimension; finally, the mass-customization two-dimensional spatiotemporal model is optimized to finish the mathematical modeling for the mass-customization two-dimensional spatiotemporal model; the heuristic optimization method is adopted for solving according to the mathematical model; the optimization of the time dimension aims at the operation process; the optimization of the spatial dimension aims at the product structure, so as to realize the purpose of delaying the decoupling point of the customer order forms. Based on the modeling and optimization of the mass-customization two-dimensional spatiotemporal model, the problem that the mass-customization two-dimensional spatiotemporal model which is a complex nonlinear multi-object optimization model cannot adopt the common mathematical programming method for solving is solved.

The invention provides a method and device for extracting inspection key component points, and belongs to the field of power detection, and the method for extracting the inspection key component points comprises the steps: S1, employing tower point clouds of different tower types as training samples, and determining the length of an insulator according to the different tower types and voltage levels; s2, performing vertical layering on the tower point cloud according to the insulator height, taking each layer of finely classified point cloud as a minimum learning unit of deep learning, performing model training on each unit to obtain a deep learning neural network model, and performing fine classification on tower point clouds of different tower types by utilizing the neural network model;s3, based on a fine classification result, carrying out monomerization on the tower point cloud of each tower by utilizing a model construction method; and S4, extracting the position of the component point from each monomeric model to serve as a photographing point position. By utilizing the method, the workload of manually selecting key component points can be reduced, and the working efficiency is improved.

The invention discloses a multi-model building method for underground water numerical simulation. The method comprises the following steps that 1, a model 1# and a model 2# are built according to original terrain data and existing measured terrain data; 2, the initial mean and the initial variance of a permeability coefficient log are generated according to the permeability coefficient data obtained through collection and an in-situ hydrogeological test, each group of samples are generated according to an AM-MCMC algorithm, each group of the samples comprise permeability coefficient log values of the corresponding grid number, each group of the log values are converted and input into the model 1# for analog computation, 100 groups of parameters containing certain data are screened according to the acceptable conditions of analog computation, corresponding output data is obtained, and then the screen 100 groups of the parameters are input into the model 2# to obtain the corresponding output data; 3, analysis is conducted according to AICc criteria, the optimal value interval of the models is analyzed by calculating the model probability or weight on the basis of the AICc criteria. Accordingly, the precision of an underground water analog model is improved, and a basic model is provided for underground water environment self-purification simulation study.

The embodiment of the invention provides a model self-adaption method and system based on an intelligent computing framework. The method comprises the following steps: based on an automatic machine learning technology, realizing automatic feature selection, hyper-parameter optimization and neural network architecture search during model construction; based on an expanded open type neural network exchange technology, realizing model cross-framework migration after model construction. According to the embodiment of the invention, important steps related to features, models, optimization and evaluation are automatically learned through a self-adaptive method during model construction; therefore, the manual intervention degree in the machine learning process is reduced, the machine learning method can be more simply and efficiently used by a user, the algorithm library of each mainstream framework is abstracted, the models are stored in a unified format by means of a neural network exchange technology, cross-framework migration of the models is realized, the reusability of the models is improved, and the development cost is reduced.

The invention discloses a user layering model construction method, a user layering model construction system, an operation analysis method and an operation analysis system. The user layering model construction method comprises the steps that: the user behavior distribution data of a process link in a long-rent apartment scene are obtained; feature data representing three dimensional indexes, namely, closeness, frequency and quota are extracted from the user behavior distribution data, wherein the feature data comprise browsing record days, daily average click search result times, daily averageaccess duration, list page daily average browsing times, detail page daily average browsing times, reservation times and attention success times; and a user layering model for user value classification is constructed according to the three dimensions of the closeness, the frequency and the quota and the feature data. According to the method, seven pieces of feature data which are suitable for a long-rent apartment scene and can serve as the three-dimensional indexes are determined firstly, the user layering model suitable for the long-rent apartment scene is constructed according to the feature data, user behavior analysis is carried out on a user according to the model, refined operation is achieved, and operation efficiency is improved.

The invention is applicable to the technical field of software tests, and provides a model construction method and device for a terminal application. The model construction method comprises the steps of obtaining status sequences from application starting to application closing of the terminal application when the terminal application is running normally, wherein the state in the status sequences comprises an operation order input by a user and an interface pack name which is skipped in the operating process; forming a behavioral training set through the obtained multiple state sequences, wherein any state of the terminal application is included in at least one state sequence of the multiple state sequences; constructing a Markov test model related to the operation of the application based on the behavioral training set. The construction procedure of the model is simple, by adding the state sequence into the behavioral training set, an original test model can be improved, thus the situation that if the functions of the application are modified, the model need to be reconstructed completely is avoided, and the model construction efficiency of the terminal application is improved.

The invention provides a model construction method and device based on a rejection inference method and electronic equipment. The method comprises the following steps: acquiring full-scale sample data, and defining positive and negative samples for the received samples to establish a training data set, wherein the the training data set comprises user feature data and financial performance data; constructing an initial risk assessment model, and training the initial risk assessment model by using the training data set; scoring the rejected samples by using the trained initial risk assessment model to obtain the deterioration probability of each rejected sample; adopting a rejection inference method to carry out weighted expansion on rejected samples, and carrying out weighted processing onall received samples; integrating the receiving sample and the rejecting sample after weighting processing, and establishing a new training data set; and retraining the initial risk assessment model by using the new training data set to obtain a final risk assessment model. According to the method, the problems of sample deviation or sample data missing and the like are effectively solved, and themodel prediction precision is improved.

The invention discloses an electrostatic dust collector three-dimensional model construction method based on CATIA. The method comprises the following steps of component classification, component parameter design, parameterized component creation, parameter table creation, component library establishment and three-dimensional model construction. The overall, rapid and linked parameter adjustment and layout design of the electrostatic dust collector model are carried out based on three-dimensional modeling software CATIA, wherein a universal electrostatic dust collector component library is formed through rapid parameterization design, all components in the library are directly called when the three-dimensional model of the electrostatic dust collector is constructed, and the modeling speed is high; by dynamically modifying model parameters, quickly endowing the model with attributes and increasing the information amount of the model, a more accurate, comprehensive and systematic information model is constructed, the efficiency and precision of modeling and parameter design can be effectively improved, and the design cycle and resource consumption are reduced. Thus, a user only needs to modify the parameter value of the parameterized part template so that the target part can be established; or only the key part of the part needs to be defined as a certain parameter, and the product is designed and optimized by modifying the parameter.

The invention discloses a gravity field three-dimensional model construction method based on a Delaunay triangulation network, and the method comprises the steps: obtaining gravity field sampling data, converting longitude and latitude coordinates in the gravity field sampling data into projection coordinates X and Y, carrying out the factor conversion of a gravity abnormal value G in the gravityfield sampling data, and obtaining a gravity conversion value Z; enabling X, Y and Z jointly to form gravitational field three-dimensional data; and then, subdividing the gravitational field three-dimensional data by using a Delaunay triangulation algorithm to obtain a gravitational field triangulation network, i.e., obtaining the gravitational field three-dimensional model for three-dimensional space analysis of the gravitational field. The method can improve the expression accuracy of gravity field data; the gravitational field three-dimensional model can be used for three-dimensional spaceanalysis of the gravitational field, and the flexibility and accuracy of gravitational field data analysis can be improved.

The invention relates to a soil quick-acting potassium analysis model construction system and method based on satellite images, belongs to the technical field of remote sensing technologies and variable fertilization, and aims to solve the problems of high randomness and high parameter adjustment uncertainty of a traditional BP neural network in soil content acquisition and monitoring. The systemcomprises an original data acquisition module, an image wave band construction module, a sensitive wave band transformation module, a transformation wave band analysis module and the like. The methodcomprises the following steps: acquiring a soil sample in a target area according to requirements of soil sample acquisition technical specifications, then preprocessing an image waveband, carrying out mathematical transformation on a sensitive waveband, finally training a neural network model in a model training verification module, and carrying out verification analysis on an obtained result; accurate and comprehensive fertilization is realized, and the cost is reduced, so that excessive cost input and environmental pollution caused by excessive fertilization are avoided, soil hardening caused by excessive fertilization is avoided, and crop growth is reasonably and efficiently promoted.

The invention provides a power system fault sample generation and model construction method for transfer learning. The method comprises the steps: obtaining an initial power system transient stabilityevaluation model; obtaining a power system fault sample; inputting the power system fault sample into an initial power system transient stability evaluation model to obtain an output result of a fullconnection layer corresponding to the initial power system transient stability evaluation model; taking an output result of the full connection layer as an input feature of a classification layer ofan initial power system transient stability evaluation model; and training a classification layer based on transient stability and transient instability samples in the power system fault samples to obtain a power system transient stability evaluation model meeting evaluation requirements. According to the method, the problems that when the operation mode and the topological structure of the powersystem are greatly changed, an original power system transient stability evaluation model is not suitable any more, and time and memory are consumed when one power system transient stability evaluation model is retrained are solved.

The invention provides a shipbuilding-oriented one-stop construction system and method for multiple industrial mechanism models. The system comprises a modeling preparation layer, a basic model construction layer, a model reconstruction layer and a model storage layer, wherein the modeling preparation layer is used for collecting and summarizing model demand characteristic information; the basic model construction layer is used for modeling a basic model; the model reconstruction layer is used for further editing and reconstructing the basic model; and the model storage layer is used for analyzing the types of the models generated by the basic model construction layer and the model reconstruction layer, and storing the models in corresponding model libraries in a classified manner according to the types of the models. According to the one-stop construction system for the multiple industrial mechanism models for shipbuilding, the multiple industrial mechanism models involved in the industry are rapidly constructed and regenerated through a one-stop basic model development and model regeneration technology, so the construction speed and quality of the models are improved, and high-quality development of shipbuilding is assisted.

The invention discloses a model building system of a concrete continuous box girder bridge. The model building system comprises the specific steps that 1, comparison and selection are conducted on a bridge type building scheme, and a bridge construction scheme is drafted; step 2, drawing up the size of the upper structure of the bridge; 3, establishing a space finite element model of the bridge structure by applying midas civil software; 4, performing curve continuous box girder counter-force checking calculation, and continuously adjusting the structure size or reinforcing bars until all checking calculations meet the standard; 5, structural internal force calculation and internal force combination are carried out; the invention relates to the technical field of bridge construction. According to the model building system of the concrete continuous box girder bridge, the problems that an existing data simulation system is too tedious in process, large in error and incomplete in analysis factor are solved.

The invention discloses a construction system of a digital twinborn body of a pile gripper. The construction system comprises a model construction module, a multi-dimensional motion state synchronousmapping module and an omnibearing visual interaction module, and the model construction module is used for carrying out modeling on the pile gripper and the offshore platform according to SolidWorks software and importing the modeling into a Unity 3D platform to obtain a motion relationship hierarchical diagram of the digital twinborn body of the pile gripper; the multi-dimensional motion state synchronous mapping module is used for obtaining the operation state data of the digital twinning body of the pile gripper in real time through a PLC control system and storing the operation state datain a PostgreSQL database; and the omnibearing visual interaction module is used for observing the operation state of the digital twinning body of the pile gripper according to the operation state dataand giving an early warning when any detection data in collision detection, distance detection and speed detection exceeds a preset value. According to the construction system of the digital twinbornbody of the pile gripper, the intelligent degree of the pile gripper is improved, and the problems that large equipment is poor in synchronism, low in accuracy and poor in interactivity are solved.

The invention provides an information prediction method and device, a terminal and a readable storage medium. The information prediction method comprises the steps of receiving user portrait information and a model construction instruction; cleaning numerical data in the user portrait information according to a model construction instruction to generate preprocessed data; performing WOE code conversion on the character data in the preprocessed data to generate conversion values corresponding to the character data, and updating the character data in the preprocessed data to the corresponding conversion values to obtain input data; performing model training according to the input data and a preset model system to generate a target model; and obtaining a to-be-predicted sample and inputting the to-be-predicted sample into the target model to generate a prediction result. The prediction result output by the model constructed by the invention is high in accuracy.

The invention provides an FPGA (field programmable gate array) interconnection line time-delay acquiring method and a system utilizing the same. The FPGA interconnection line time-delay acquiring method includes: the model building and analyzing step: dividing an interconnection line of the FPGA into a plurality of models, determining the number of varied paths caused by the number of varied loads of each model; the initial processing step: obtaining time delay of each path by a net list extracted by a layout according to variation of the number of loads and filling the time delay parameters of the models into a database; the time delay processing step: obtaining interconnection line models by searching the database during layout arranging and wiring, then calling the time delay parameters of the corresponding model to acquire the total time delay of the integral interconnection line by means of fitting of numerical values.

The invention provides an analysis method, which comprises the steps of determining an analysis time period in response to a received analysis instruction for a target object, the analysis time periodbeing related to a timestamp of predetermined time series data of the target object; obtaining first time series data of the target object in the analysis time period; obtaining a first time sequencefeature set based on the first time sequence data; at least inputting the first time sequence feature set into a pre-established first analysis sub-model to obtain a first intermediate result; and sending the first intermediate result to the joint platform device, so as to enable the joint platform device to obtain an analysis result based on the first intermediate result and a second intermediate result from a second mechanism server, wherein the second mechanism is a mechanism for joint analysis with the first mechanism. The invention further provides a model construction method, an analysis device, a model construction device, a joint analysis system, electronic equipment and a computer readable storage medium.