24 results about "Point model" patented technology

The invention discloses a product model detection method and system based on welding spots. The method comprises the steps that a welding spot identification mark is obtained, a welding spot file matched with the welding spot identification mark is identified from a product model, and the welding spot file is a data file comprising a welding spot model; identifying a part model having an interference relationship with the welding spot model based on the position relationship of the welding spot file in the product model; taking an intersection surface obtained by intersecting the part models as a welding surface of the part models, projecting the welding spot model onto the welding surface to obtain a welding spot projection position, and detecting a shortest spot edge distance from the welding spot projection position to a welding surface boundary; based on the shortest point edge distance, the width of the welding face is determined through an auxiliary face intersection method; and detecting whether the product model is qualified or not based on the shortest point edge distance and / or the width of the welding surface. The technical problems that in the prior art, welding spot detection mainly depends on manual estimation, so that subjectivity is high, efficiency is low, and missing detection is prone to occurring are solved.

The invention provides a train integrity detection method and system for constructing a dynamic envelope based on a multi-mass-point model, and belongs to the technical field of rail transit train operation safety detection. Estimating the real-time length of the baseline by using Kalman filtering; pID control is adopted to analyze and calculate the telescopic displacement of the carriages in the train running process; and comparing the calculated baseline length with the train safety envelope, and judging whether the integrity state of the train is normal or not according to a comparison result. According to the invention, the train length is detected by using the double-difference carrier phase; a train dynamic safety envelope is constructed, a train multi-mass-point model is constructed by researching longitudinal stress analysis such as resistance and workshop acting force of a high-speed train, and compartment extrusion displacement in the train running process is analyzed and calculated through PID control. A train dynamic safety envelope is formed by the data, communication delay and a redundancy threshold value and serves as a judgment basis of train integrity detection, and the efficiency and the real-time performance of train integrity detection are improved.

The invention provides an MPM-based three-dimensional slope deformation analysis method and system, and the method comprises the steps: firstly obtaining the spatial geometric structure data of a slope region, constructing a three-dimensional material point model based on the spatial geometric structure data, discretizing a slope material into a material point set containing position coordinates and physical attribute information, and carrying out the discretization of the material point set; establishing a background grid system covering the material point set, then performing interaction calculation on the material point set and the background grid system, solving a kinetic equation on background grid nodes to obtain motion state parameters, mapping the motion state parameters back to material point updating information, generating a material point state set containing deformation characteristics, and finally performing deformation calculation on the material point state set. And finally, based on the position coordinate variable quantity and the physical attribute change parameters in the material point state set, analyzing the deformation trend and the potential instability area of the three-dimensional slope, and generating a slope deformation analysis report containing deformation gradient distribution and a stress concentration area, so that the geometric and stratum information of the slope can be comprehensively considered, and the deformation process can be accurately simulated.

This invention discloses an industrial Internet of Things (IIoT) platform, relating to the field of IIoT technology, comprising a point modeling module, a business analysis module, a point matching module, a state chain generation module, a state chain operation module, a topic scheduling module, a linkage action orchestration module, and an execution receipt verification module. The point modeling module generates point information, point reliability, and on-site message data; the business analysis module generates a business requirement diagram; the point matching module generates point binding relationships; the state chain generation module generates a state chain; the state chain operation module calculates state satisfaction values ​​and valid state satisfaction values, and generates state correction results; the topic scheduling module determines message transmission strategies; the linkage action orchestration module generates on-site control commands; and the execution receipt verification module confirms the execution status and generates a receipt correction action. This invention improves the consistency and traceability of business state transitions, anomaly correction, and action execution confirmation.

A steel structure digital restoration device and method for digital twin construction, the device comprises a structure recognition module, the structure recognition module is configured to: the point cloud information of the acquired steel structure is executed in a virtual three-dimensional space according to its embedded space coordinate information based on the corresponding restoration operation of position, the total model field about the distribution of point cloud in space is acquired, and the restored point cloud information is in the form of a point model in the total model field;Based on the distribution of point cloud, the position with the best point model density is automatically selected, slicing is carried out to obtain the basic cross-section data of the structure;Based on the image in the basic cross-section data, the identification and judgment of the type of steel structure are executed, after correctly identifying the type of steel structure, the three-dimensional model fitting restoration of slicing is executed based on the identified type of steel structure.The present application can directly restore the three-dimensional pipeline model with high precision, and the engineering personnel can obtain the information of all steel structures in the whole factory by directly observing the three-dimensional steel structure model with sufficient rich details.

The invention discloses a large-span truss structure digital twinning establishment method based on a three-dimensional laser point cloud, and the method comprises the steps: firstly, carrying out the modular segmentation of point cloud data based on structural features in a data preparation stage, and dividing the point cloud data into modules which accord with the size specification; secondly, in the skeleton point recognition process, a multi-level division method based on spatial point cloud features is adopted, point cloud clusters are refined step by step, and the centroids of the point cloud clusters are extracted; then, performing automatic skeleton point identification based on the standard skeleton point model, and generating the standard skeleton point model through truss data topology; and finally, performing system reconstruction of a digital twinborn model, accurately determining connection nodes of the components and reconstructing a truss structure module by calculating directions and positions of framework points of the components, and assembling to form a complete truss entity structure digital twinborn. Point cloud data of an actual structure is efficiently, accurately and automatically processed, and a digital twin model capable of truly reflecting the deformation state of the structure is established.

The application provides a three-dimensional terrain geological modeling method and device based on material point method simulation, and relates to the technical field of geotechnical engineering numerical simulation, and comprises the following steps: S1, extracting elevation data of a three-dimensional terrain geological region model; S2, converting the elevation data into three different type model component data and processing; S3, based on the generated model component data, discretizing the model into material points to generate a material point model; S4, importing the generated material point model into a material point method calculation program to perform material point method simulation, and obtaining large deformation of the geology according to the material point method simulation; the application effectively solves the problem of large deformation simulation modeling caused by complex and diverse terrain and landform, is efficient, stable, simple and feasible, significantly improves the modeling efficiency of the material point method and the quality of the generated material point model, and effectively promotes the application of the material point method simulation in the field of geotechnological engineering numerical simulation.

The present invention relates to the field of artificial intelligence technology, and discloses a target detection optimization method and system based on a CenterPoint model. The method comprises obtaining an original image of a current frame; inputting the original image into a CenterPoint model for inference to obtain target category, center offset position, and target size information; inputting the target category, center offset position, target size information, and a historical target center position set stored in the previous frame into a pre-trained target center point motion model to obtain a target center position set; rendering the target category, target center position set, and target size information into the original image to obtain a rendered image; performing target detection based on the rendered image, and tracking the target based on the target detection result. The method has the following effects: achieving accurate detection of fast-moving or obscured targets to improve the accuracy of target tracking.

This invention discloses a method, apparatus, and storage medium for establishing ACM weld point models. It automatically detects and deletes target weld points through scripts or program instructions, determines the size of the weld nugget refinement mesh based on the set weld nugget diameter, obtains modeling information and corresponding finite element models of non-coincident weld points, and automatically locates and adaptively updates the refinement area of ​​the connector based on the weld nugget refinement mesh and the modeling information of non-coincident weld points. It traverses all target weld points to be updated, completing the refinement of the weld point connector mesh and the batch automatic generation of new ACM weld points. This method ensures the accuracy of the ACM weld point model and improves modeling efficiency.

The present invention is applicable to the technical field of machine vision, and in particular relates to a training analysis method for a generative adversarial network. The method comprises: establishing an isolated point model; analyzing behaviors near real points in a real sample distribution by means of the isolated point model, to determine whether a generative adversarial network training process is locally stable; analyzing an incorrect local minimum value and an approximate mode collapse during the generative adversarial network training process by means of the isolated point model; analyzing, by means of the isolated point model, the situation that occurs for generated points isolated from the real points; and analyzing the effect of a kernel width during the generative adversarial network training process by means of the isolated point model. By means of the isolated point model, the training dynamics of a kernel in a generative adversarial network that jointly updates a generator and a discriminator can be decoupled, the stable local balance and the incorrect local minimum value are analyzed, and the significant effect of the kernel width during the generative adversarial network joint training is analyzed.

The present invention discloses a method for identifying the parameters of an integral time-delay zero-point model and rolling simulation of water levels, which relates to the technical field of power station optimal dispatching. The method includes: constructing a one-dimensional river hydrodynamic model based on the Saint-Venant equations; wherein, the one-dimensional river hydrodynamic model is used for calibrating the Manning coefficient of the model, designing working conditions, and obtaining the stable operation state of the working conditions; when the one-dimensional river hydrodynamic model is in a stable operation state, a flow step is applied to the stable operation hydrodynamic model for parameter identification to obtain the key parameters under different working conditions; using a multivariable fitting method to fit each key parameter to determine the parameter function relationship; based on the parameter function relationship, introducing the time-domain expression of the integral time-delay zero-point model to calculate the water level difference between adjacent moments; and performing rolling simulation of the river water level according to the set upstream and downstream flows, the initial river water level, and the water level difference between adjacent moments. The present invention can improve the efficiency and accuracy of real-time dispatching of power stations.

The present invention discloses a method for generating a material point method model of a soil-rock mixed slope, which is used to simulate the mechanical behavior of a soil-rock mixed slope. The method comprises the following steps: first, randomly dividing the slope area to generate polygonal units that meet the target area and shape distribution; then, based on the shape contours of the polygonal units, generating random block contours through polar coordinate sampling and Fourier series fitting, and controlling the size of the block by adjusting the area ratio; finally, mapping the generated block contours to a material point array, calibrating the material points belonging to the block, and completing the pre-processing of the material point method calculation. The present invention can truly reflect the size, shape, and spatial distribution characteristics of the blocks in the soil-rock mixed slope, effectively improve the efficiency of block placement, avoid overlapping problems, and overcome the defects of the existing technology in the simplification of block shape and insufficient randomness of distribution. At the same time, it is suitable for the numerical simulation of large deformation problems, has high calculation accuracy and broad engineering application prospects.

The invention discloses a non-intrusive user load monitoring method and system for a sequence-to-point model, and relates to the field of non-intrusive power load monitoring, and the method comprises the steps: constructing a load decomposition network structure model, completing the pre-training of a decomposition network model through the supervised training and verification of a public data set, and obtaining a non-intrusive user load model; the transfer learning of the model is completed through supervised training of a private data set, real-time feedback or adjustment is provided according to a monitoring result, and the performance of the model in practical application is evaluated and optimized. According to the method, the required data set is constructed in a mode of combining the public data set and the private data set, standardization processing is performed to solve the problem of insufficient data sample size faced by a deep learning model NILM, transfer learning of the model is performed in combination with data in the private data set, so that the method adapts to different environments and requirements, and the method is suitable for large-scale popularization and application. And the generalization ability and accuracy of the model are improved, so that the problem of insufficient training data is relieved.

The invention provides a three-dimensional terrain geological modeling method and device based on material point method simulation, and relates to the technical field of geotechnical engineering numerical simulation, and the method comprises the following steps: S1, extracting elevation data of a three-dimensional terrain geological region model; s2, converting the elevation data into three different types of model component data and processing the data; s3, based on the generated model component data, dispersing the model into material points, and generating a material point model; s4, importing the generated material point model into a material point method calculation program to carry out material point method simulation, and obtaining geological large deformation according to the material point method simulation; according to the method, the problem of large deformation simulation modeling caused by complex and diversified terrains and landforms is effectively solved, the method is efficient, stable, simple, convenient and feasible, the modeling efficiency of the substance point method and the quality of the generated substance point model are remarkably improved, and application of substance point method simulation in the field of geotechnical engineering numerical simulation is effectively promoted.

The invention provides a c code and verilog code generation method and device, and relates to the technical field of software development and FPGA board card debugging. When the method is executed, firstly, a model is built through GCKontrol modeling software, secondly, floating-point simulation is conducted on the model, thirdly, subsystem packaging is conducted on a module, needing to automatically generate the verilog code, in the model, then the module, needing to automatically generate the verilog code, in a subsystem is calibrated according to a floating-point simulation result, then, the calibrated model is simulated, and finally, the model is subjected to automatic generation of the verilog code. And finally, if the simulation result of the model accords with the expectation, generating a c code and a verilog code at the same time by utilizing the calibrated model. In this way, the c code and the verilog code are automatically generated at the same time.

The invention relates to the technical field of electric power system inspection, and provides an inspection model debugging configuration method and system, and the method comprises the steps: splitting a complex inspection task into a plurality of independent and correlated data models, and carrying out the configuration, comprising a navigation map model, a scene model, a point position model, a robot model, an electronic map model, a stop position model, a preset position model and the like, the data models with different functions are decoupled, the low coupling degree between data is achieved, when a certain function is modified or added, the whole architecture does not need to be comprehensively adjusted, and the cost is reduced. According to the method, the development cost and risk are reduced, each model can be independently updated and optimized without influencing other models, and the maintainability and expandability of the whole model are greatly improved.

The invention discloses a data accelerated reasoning method and device based on ONNX and a medium. The method comprises the steps that a quantized acceleration fixed-point model is adopted to conduct acceleration reasoning on to-be-processed data, a data reasoning result is obtained, and the generation process of the acceleration fixed-point model is as follows: an original ONNX floating-point model is quantized, and a quantized IR fixed-point model is generated; replacing a user-defined operator in the IR fixed-point model with an ONNX native operator to generate an accelerated fixed-point model; and performing precision verification on the acceleration fixed-point model by adopting the verification data set, and determining a final acceleration fixed-point model under the condition that the precision is met.

The invention discloses an incremental non-intrusive decomposition method and system considering load dynamic change, and relates to the technical field of power load management analysis, and the method comprises the steps: receiving a public data set, determining training related information based on the public data set, obtaining power utilization data and total electricity meter data of to-be-trained equipment based on the training related information, and carrying out the training of the to-be-trained equipment; inputting the power consumption data of the to-be-trained equipment and the total ammeter data into a pre-established sequence arrival point model, and training to obtain an independent decomposition model of the to-be-trained equipment; performing traversal detection on the to-be-trained equipment to obtain newly-added electric equipment, obtaining data of the newly-added electric equipment, inputting the data of the newly-added electric equipment into an independent decomposition model of the to-be-trained equipment by using a transfer learning technology, and performing fine tuning to obtain a fine-tuned decomposition model; and inputting the total ammeter data of the to-be-trained equipment into the fine-tuned decomposition model to obtain the power consumption data of the newly-added power consumption equipment.

The invention discloses an anti-fraud method and system, which is based on a feature importance distribution mechanism of a gradient boosting decision tree (GBDT), automatically learns an optimal weight combination of relation features and individual features, and identifies a high-order interaction effect of the relation features of gang and individual features of a single point by using a nonlinear fitting capability of a tree model. A graph algorithm captures a relation change relation and supplements the relation change relation to traditional individual change features, the problems that a traditional single-point model cannot capture an incidence relation and an independent graph algorithm ignores individual behaviors to cause misinformation are solved, and fusion of multi-dimensional fraud features is achieved. The individual features and the relation features are combined and input into a CatBoost algorithm for reasoning and judgment, and the higher accuracy rate and the higher recall rate compared with a single graph algorithm model and a single Boosting algorithm model are obtained.

The present invention is applicable to the technical field of machine vision, and in particular relates to a training analysis method for a generative adversarial network. The method comprises: establishing an isolated point model; analyzing behaviors near real points in a real sample distribution by means of the isolated point model, to determine whether a generative adversarial network training process is locally stable; analyzing an incorrect local minimum value and an approximate mode collapse during the generative adversarial network training process by means of the isolated point model; analyzing, by means of the isolated point model, the situation that generated points isolated from the real points occurs; and analyzing the effect of a kernel width during the generative adversarial network training process by means of the isolated point model. By means of the isolated point model, the training dynamics of a kernel in a generative adversarial network that jointly updates a generator and a discriminator can be decoupled, the stable local balance and the incorrect local minimum value are analyzed, and the significant effect of the kernel width during the generative adversarial network joint training is analyzed.

The invention relates to the field of elevation point modeling information management in constructional engineering, in particular to an elevation point attribute statistics and reverse search method and device for constructional engineering, electronic equipment and a storage medium. Obtaining target elevation point attributes from an attribute database; on an interactive interface of the modeling tool, attributes are displayed through elevation point attribute statistics and reverse checking positioning. According to the method, the elevation point attribute can be checked as required without selecting specific elevation points one by one, the problem of low elevation point information management efficiency is solved, the manual workload is reduced, and the use convenience and design efficiency of modeling software are improved.

Embodiments of the present application provide a method for processing motion data during kicking and a training evaluation method for football special training. The method includes: based on n sets of motion data collected by sensors installed on a football, using a 3D engine to generate a spatial coordinate point model for each set of motion data; establishing a core data region according to core spatial points and other spatial points, using the mini-batch gradient descent method to determine an objective function, performing convolution calculation on the motion data and the objective function to determine whether a certain moment is a kicking action, and calculating the kicking moment, the number of kicks, the kicking direction and the kicking force accordingly. After importing them into the training evaluation system, the special score and the overall score of football training are obtained, and a comprehensive evaluation five-dimensional graph is generated. The present application can obtain the special data of football special training more accurately and quickly in real time, and can comprehensively evaluate various indicators in the football training process, assist relevant personnel in judging the situation of football training, and make improvements accordingly.

The invention discloses an anti-loose fuse forming process technological parameter simulation analysis method. The method comprises the following steps that an initial fuse model, a bolt fixing point model and a rotation auxiliary model are assembled to obtain a double-helix fuse forming simulation model; the fuse forming process is decomposed into three analysis steps of folding, rotating and pulling; dividing grids for the initial fuse model, the bolt fixing point model and the rotation auxiliary model, and setting a load mode for the three analysis steps to perform simulation analysis; acquiring limit working loads corresponding to different process parameters according to a fuse failure mode criterion; and performing an orthogonal simulation test on each group of process parameters, and taking the group with the maximum limit working load as an optimal process parameter combination. By fully simulating the process characteristics of the fuse forming process, initial prestress and deformation are established for the loaded process, more accurate initial boundary conditions are provided, and the accuracy of the simulation result is improved.

The present application belongs to the technical field of three-dimensional geological modeling, in particular relates to a kind of GIS-MPM seamless integration dynamic three-dimensional geological model construction method, this method based on GIS platform secondary development seamlessly integrates GIS with MPM (material point method), utilizes GIS to carry out three-dimensional modeling and three-dimensional visualization, utilizes MPM to carry out geotechnical mass large deformation numerical calculation, simultaneously, it is combined with the mutual conversion technology between elevation grid data model and material point model, and the adaptive technology of material point and background grid, constructs the dynamic three-dimensional geological model that can carry out quantitative display.This method makes up the deficiency in traditional static modeling method, provides effective way for improving three-dimensional geological model quantitative calculation analysis and dynamic simulation display capability, makes three-dimensional geological model truly have "vitality", therefore have certain practical significance and application value.