31 results about "Dual graph" patented technology

The invention discloses a three-dimensional physical field real-time prediction method and system based on geometric deep learning and physical constraint, and relates to the technical field of computer-aided engineering and artificial intelligence. The method comprises the following steps: directly extracting native boundary representation data (B-Rep) of a three-dimensional model from a computer aided design system; constructing a heterogeneous dual graph taking a parameterized curved surface as a graph node, skipping finite element grid division, and aggregating local and global topological features by using a graph neural network; in combination with a physical information driving mechanism, a partial differential equation (PDE) residual error is introduced as a loss function for constraint training, and generalization prediction of a novel geometric structure is realized; and finally, the physical field state quantity is predicted through direct regression and is rendered in real time. An incremental reasoning mechanism based on a local topology subgraph is adopted, millisecond-level physical field real-time feedback under design modification is achieved, and the method is suitable for scheme rapid screening and trend prediction in the initial stage of design.

The invention discloses a logical reasoning method and device based on knowledge enhancement and capsule network collaboration and a medium. The method comprises the steps that a public knowledge local database is constructed, and a vectorization database and a graph network database are deployed locally; performing preliminary retrieval according to the logical reasoning text, and expanding a retrieval range based on a graph network; inputting the logic text, the problem and the related background knowledge into a large language model to generate a logic reasoning link, and integrating all contents to obtain a logic reasoning text after knowledge enhancement; constructing a semantic graph and a connection graph according to the logical reasoning text after knowledge enhancement; the semantic graph and the connection graph are respectively imported into a capsule network, and a dynamic routing algorithm is applied to promote low-level capsules to be progressive to high-level capsules; respectively executing an average fusion operation on the plurality of high-level capsules in the double images to extract global features of the images; global features of the double graphs are integrated through a fusion strategy, final knowledge text features are formed, and final options are obtained.

The invention belongs to the technical field of pipeline safety assessment, and discloses a multi-dimensional information interconnection and autonomous evolution collaborative pipeline full-state safety assessment method. And capturing a high-order relationship of data through double hypergraph reasoning of the instance-level hypergraph and the modal-level hypergraph to realize efficient interconnection. According to the method, mode-level and instance-level hypergraph information features are extracted through hypergraph information propagation, high-order correlation is mined through double-graph information aggregation, cross-mode and cross-instance consistency information and exclusive information are output after feature recombination, multi-dimensional data deep fusion is promoted, and high-quality data support is provided for follow-up pipeline full-state safety assessment. A two-stage autonomous evolution mechanism of intra-class progressive calibration and inter-class knowledge migration is respectively adapted to slight fluctuation and significant change scenes of the deep sea environment: precise adaptation of environment perturbation is realized through dual-branch feature extraction and dynamic weight adjustment in a domain; model parameter dynamic optimization is completed between domains through spatial-temporal feature clustering and cross-domain knowledge migration, and dynamic environment self-adaption can be achieved without manual intervention.

The invention discloses a space-time dual graph representation-based staying hot spot sequence inference method, which comprises the following steps of: constructing a space-time staying hot spot transfer relation graph based on a waybill track and staying hot spot data, and modeling a relation between a staying hot spot and a driving path; constructing a dual graph coding method based on a graph convolution variational auto-encoder model, and introducing an attention mechanism to jointly learn a space-time dynamic relationship between a staying hot spot and a driving path; performing embedding generation on the query request by using a denoising diffusion model, and capturing correlation between a historical transportation route and a current query; on the basis of a decoder of a recurrent neural network, a mask mechanism is introduced, an unreasonable staying hotspot sequence is effectively pruned, and the inference precision and efficiency are improved; performing model optimization by adopting multi-stage joint loss, and performing training by using an Adam optimizer; and deducing a staying hotspot sequence conforming to the driver behavior mode based on the historical track and the current query request. The method has wide application value.

The present invention discloses a method for predicting the popularity of event propagation based on a graph neural network. For each event text in an event propagation data sample, the feature vector of each user in the social network in the propagation of the event text is determined, and the temporal feature cascade graph corresponding to the event text is determined. A dual graph neural network is used to obtain the temporal user feature sequence of the event text based on the user feature vector and the temporal feature cascade graph. An event propagation popularity prediction model including a temporal convolutional neural network and a summation and pooling module is constructed. The event propagation popularity prediction model is trained using the temporal user feature sequence and popularity value of the event text in the event propagation data sample. The popularity value of the event currently propagating in the social network is predicted using the event propagation popularity prediction model. The present invention combines a graph neural network and a temporal convolutional neural network to improve the performance of event propagation popularity prediction.

The invention relates to a structure-semantic double-graph collaborative SAR image quality evaluation method, which comprises the following steps of separating an SAR image, and constructing a double-channel quality measurement set for independent modeling; indexes are obtained, correlation calculation is carried out, an index map is constructed, and each node corresponds to one index; on the basis of the index map, the dependency relationship between indexes is calculated from the structure dimension and the semantic dimension through self-supervised learning, the dependency relationship of the two dimensions is integrated, and an image level score based on the integration result is output; and constructing a measurement index shared with system analysis and a measurement index special for information retrieval, and carrying out quality measurement evaluation: after extracting a structure sub-graph, integrating an IR special index which is not contained in SAR prior into a global graph. The method is superior to a reference model in the aspects of recognition accuracy, recall rate, precision and efficiency, and a solution with high universality and good interpretability is provided for cross-modal image quality evaluation.

The present invention relates to a small-sample bearing fault diagnosis method based on a dual-graph network. The method involves selecting a test sample and a training sample of bearing faults, preprocessing them, collecting data points, and constructing a metadata set. A convolutional neural network is then used to extract features from the metadata set. The extracted features and the annotation information of the training samples are then input into the dual-graph network to determine the fault type of the test sample. Compared with existing technologies, this method offers advantages such as high accuracy and wide applicability.

The invention provides an interest point recommendation method and device based on an association mode and heterogeneous double graphs and a storage medium. The method comprises the steps that a user information set, an interest point information set, an interest point category information set and a user sign-in time information set are acquired; constructing an association mode according to the user information set, the interest point information set, the interest point category information set and the user sign-in time information set; respectively constructing a first heterogeneous graph and a second heterogeneous graph according to the association mode and the set; according to the association mode, based on vector embedding, respectively determining a first decomposition graph of the first heterogeneous graph and a second decomposition graph of the second heterogeneous graph; based on an information entropy weighting method, determining a first information entropy and a first weight corresponding to the first heterogeneous graph, and determining a second information entropy and a second weight corresponding to the second heterogeneous graph; and according to the first information entropy, the first weight, the second information entropy and the second weight, determining a recommendation score of each interest point. According to the method, the interest point recommendation accuracy can be improved.

The embodiment of the invention provides a maximum segmentation solving method based on primal dual graph neural network learning optimization and related equipment, and the method comprises the steps: firstly, obtaining an image maximum segmentation model which comprises an adjacent objective function and a one-hot vector parameter; then, simplex relaxation conversion is carried out on one-hot vector parameters in the image maximum segmentation model, a relaxation continuous optimization model is obtained, and the relaxation continuous optimization model comprises continuous parameters; thirdly, performing iterative solution on the relaxation continuous optimization model based on an adjacent objective function, continuous parameters and a dual hybrid gradient neural network to obtain a relaxation feasible solution; and finally, sampling is carried out based on the probability distribution of the relaxation feasible solution, a target feasible solution of the image maximum segmentation model is obtained, the target feasible solution is used for segmenting image data corresponding to the image maximum segmentation model into multiple pieces of subset image data with the maximum total weight, and the solving efficiency of the maximum segmentation model is greatly improved.

This invention relates to the field of software engineering technology and discloses a method and system for generating code annotations based on a dual graph neural network. The method includes: acquiring source code and natural language annotations as source code, preprocessing the source code, and constructing training data based on the preprocessed source code and natural language annotations; constructing a code processing model, training the code processing model using the training data, and obtaining a trained code processing model. The code processing model includes a dual graph module, an encoder, an aggregator, and a decoder. The dual graph module parses the preprocessed source code to obtain an abstract syntax tree (AST), and constructs a syntactic dependency graph and a semantic dependency graph based on the AST; the encoder obtains a first encoding representation based on the syntactic dependency graph and a second encoding representation based on the semantic dependency graph; the aggregator obtains an aggregate graph representation based on the first and second encoding representations; the decoder generates natural language annotations based on the first, second, and aggregate graph representations; and inputting the code to be annotated into the trained code processing model to obtain natural language annotations. This invention can more effectively encode code structure, improve the accuracy of annotation generation, and enhance the generalization ability of the code processing model.

The present invention discloses a knowledge tracing modeling method and system based on a dual graph neural network, mainly related to the fields of intelligent education and educational big data mining. It includes a question-answering interaction node: based on the question information, adding the possible answering situation of the student to this question (i.e., answering correctly or wrongly), and fusing it into the student's question-answering interaction, which serves as the node for constructing a graph; a concept association graph: constructing a hypergraph between the question-answering interaction nodes, and using a hypergraph convolutional network to obtain the representation of each question-answering interaction node, which incorporates the association information between the questions and knowledge concepts; a directed transfer graph: constructing a directed graph between the question-answering interaction nodes, and using a directed graph neural network to obtain the representation of each question-answering interaction node, which incorporates the transfer information between the question-answering interaction nodes. The beneficial effects of the present invention are as follows: it can improve the quality of teaching work and achieve personalized teaching.

The invention relates to the technical field of data analysis, and provides a dynamic hierarchical rendering and context awareness-based report header management method, which comprises the following steps of: acquiring financial index historical data, identifying a supervision period type, processing time sequence data through a supervision period embedding layer and a three-layer attention mechanism of a hierarchical time sequence encoder, and performing dynamic hierarchical rendering and context awareness; outputting a time sequence weight feature vector; constructing a main graph and a constraint graph by using a dual graph convolutional network, propagating information through a constraint perception message transfer function, adjusting a graph topological structure according to a constraint default degree, and outputting an index relation weight matrix; inputting the time sequence weight feature vector and the index relation weight matrix into a neural symbol fusion network to obtain a final display weight value; and sorting and layering the financial indexes by adopting a dynamic layering strategy based on the weight values, and outputting a layered rendered report header. According to the method, the accuracy and adaptability of business index importance weight prediction are improved, and intelligent header layout optimization is realized through a dynamic layering strategy.

This invention provides a method and system for crack recognition in ground-penetrating radar (GPR) images based on frequency-domain enhanced YOLOv11. The method comprises the following steps: S1, constructing a GPR B-scan image dataset containing road cracks and dividing it into training and validation sets; S2, constructing a PFAE-YOLOv11 feature extraction model, which embeds a pyramid frequency attention extraction module into the YOLOv11 backbone network; S3, training the PFAE-YOLOv11 feature extraction model using the training and validation sets to optimize network parameters; S4, inputting the GPR image to be identified into the trained model and outputting initial detection bounding boxes for the cracks; S5, constructing a dual graph convolutional inference module, using the initial detection bounding boxes output in step S4 as graph nodes, and using a graph convolutional network to infer the topological relationships of the cracks, outputting corrected continuous crack paths. This invention significantly improves the accuracy, completeness, and anti-interference capability of crack detection in GPR images.

The invention provides a grid parameterization method and device, equipment, a storage medium and a program product. The method comprises the steps of obtaining a first target grid; the first target grid is a topological structure to be parameterized; the first target grid comprises a plurality of patches; segmenting the first target grid based on the dual graph to obtain a second target grid; determining whether the second target grid meets a preset condition or not; the preset condition is used for indicating whether the second target grid still needs to be subjected to secondary segmentation; if the second target grid meets the preset condition, determining the torsion resistance of the second target grid, and determining a third target grid based on the torsion resistance of the second target grid; adopting a preset parameterization algorithm to carry out parameterization mapping on the third target grid so as to obtain initial parameterization coordinates corresponding to the third target grid; and arranging the initial parameterized coordinates corresponding to the third target grid in a preset parameterized space to obtain target parameterized coordinates corresponding to the third target grid.

The present application relates to the technical field of computer-aided design, in particular to a distribution box thermal fault prediction system based on a mechanism model, which comprises a model voxelization processing module, used for calling a three-dimensional model of a distribution box containing busbars and circuit breaker components, performing spatial discretization on the three-dimensional model of the distribution box, establishing thermal conduction connection edges between adjacent voxel units, and generating a voxelized dual graph.In the present application, the number of non-empty grids is counted and the streamline fractal dimension is calculated by covering different scale space grids, so that the geometric distribution characteristics of the flow field are converted into quantifiable complexity indicators, and then the flow field dimension attenuation coefficient distribution is generated in combination with the reference dimension value under the standard operating condition, thereby realizing quantitative identification of the dust deposition state of the air inlet.The thermal fault prediction is extended from a single temperature criterion to the structure coupling and flow evolution level, the forward-looking of thermal anomaly identification is improved, and the calculation burden caused by the model size is reduced.

The present invention relates to a perspective-level text sentiment classification system based on a dual graph convolutional neural network, including a text preprocessing module for performing feature extraction on perspective-level text; a text semantic information acquisition module for capturing bidirectional semantic dependency relationships of the text; an attention encoding module for capturing the global internal correlation of the text word sequence and generating a text semantic relationship graph; a related semantic graph convolutional neural network module that applies GCN to the text semantic graph to model the sentence structure; a text syntactic information acquisition module for capturing text information based on dependency syntax; a dependency syntax graph convolutional neural network module that directly applies GCN to the sentence dependency tree to model the sentence structure; a bidirectional mapping module for exchanging relevant features between semantic GCN and syntactic GCN information; and an emotion category output module that uses a classification function to obtain the final sentiment classification result.

The present application belongs to the technical field of pipeline safety assessment, and discloses a pipeline full-state safety assessment method based on multi-dimensional information interconnection and autonomous evolution coordination. The high-order relationship of data is captured through dual hypergraph reasoning of instance-level hypergraph and modal-level hypergraph, and efficient interconnection is realized. The modal-level and instance-level hypergraph information features are extracted through hypergraph information propagation, the high-order correlation is mined through dual graph information aggregation, and the cross-modal and cross-instance consistent information and exclusive information are output after feature reorganization, which promotes the deep fusion of multi-dimensional data and provides high-quality data support for subsequent pipeline full-state safety assessment. The two-level autonomous evolution mechanism of intra-class gradual calibration and inter-class knowledge transfer is respectively adapted to the scenes of slight fluctuation and significant change of deep sea environment: within the domain, the dual-branch feature extraction and dynamic weight adjustment are realized to accurately adapt to the environmental perturbation; between the domains, the spatiotemporal feature clustering and cross-domain knowledge transfer are completed to realize the dynamic optimization of model parameters without manual intervention, and the dynamic environment self-adaptation is realized.

This paper discloses a reference image segmentation method based on cross-modal dual graph alignment, belonging to the field of multimodal image segmentation. This method innovatively proposes a "part-unity-whole" paradigm, mapping the extracted visual and textual features to a unified latent representation structure before performing cross-modal fusion. This facilitates the explicit alignment information extracted by the model, effectively enhancing the final segmentation effect.

The present application belongs to the technical field of defect detection, and discloses a kind of defect detection method, device, equipment and storage medium, comprising: mesh product image is identified to mesh, and the identified mesh is used as node to build directed graph;The dual of directed graph is calculated, and the dual graph is obtained;With the edge in dual graph as reference, set the region of interest, segment product image, obtain edge image set;Each image in edge image set is scored by the classification model obtained by pre-training, and the scoring result is pulled back to the dual graph, to obtain weighted directed graph;According to weighted directed graph, defect classification is carried out.The present application indirectly locates the edge of mesh in mesh product image through dual graph, generates edge image set, scores each edge image in edge image set through classification model, obtains weighted directed graph, and then classifies defects through weighted directed graph, to ensure that mesh product can be automatically detected for defects, and the robustness and accuracy of mesh product defect detection are improved.

A recommendation system based on double graph representation learning pre-training model, comprising: a data entry module, a meta feature extraction module, a graph construction module and an intelligent recommendation module, wherein: the data entry module enters the model data and dataset data of the external data source and outputs to the meta feature extraction module after parsing; the meta feature extraction module extracts and aggregates the meta feature vectors of the model and the dataset; the graph construction module constructs the double graph representation of the model graph and the dataset graph based on the meta feature vectors, and calculates the architecture feature similarity between the models and analyzes the label weight relationship between the datasets; the intelligent recommendation module uses a deep recommendation model based on residual graph convolution and multilayer perceptron to predict the regression accuracy of the dataset based on the double graph representation and the corresponding meta feature vectors, and then obtains the result list recommended by the model. Through meta feature extraction of the model and the dataset, and double graph representation learning based on the similarity relationship, the present application can effectively capture the complex relationship between the dataset and the model, thereby realizing efficient and accurate model recommendation.

The invention provides a graph structure-based traditional block property unit spatial relationship expression and topology construction method and system, and the method comprises the steps: firstly obtaining a property unit set in a block range, calculating the mass center of each property unit, constructing a mass center point set, and then constructing an adjacency relationship between the property units through a graph structure, the original graph is adjacent to the original graph, and a space unit defined by the boundary of the property unit is identified through plane embedding; then, establishing a dual graph by taking a space unit as a node, and completely expressing a connected network of a medium space in the block; through cooperative representation of an original adjacent graph and a dual graph, spatial organization characteristics of a traditional block can be systematically and accurately described, and a unified data basis and an analysis framework are provided for applications such as city design, spatial form analysis and fire evacuation.

The invention belongs to the technical field of road network information mining, and discloses a road network-track cooperative intersection type multimode integrated discrimination method, which comprises the following steps: identifying an undetermined intersection based on road network data, and determining a real intersection and a range thereof according to the characteristics of the undetermined intersection. The method comprises the following steps: preliminarily judging each real intersection type as a high-confidence plane intersection, a high-confidence three-dimensional intersection and an intersection to be classified, marking, constructing a dual graph of an intersection road network, and endowing node features to quantitatively express a graph structure; a to-be-classified intersection type discrimination model is constructed by using a graph convolutional network, and a high-confidence plane intersection and a high-confidence three-dimensional intersection road network graph structure are used as training data to obtain an optimization model through training. And finally, executing an intersection type discrimination task of the to-be-classified intersection road network data by using the trained model. According to the invention, the intersection type can be discriminated efficiently and accurately.

The application provides a point-of-interest recommendation method and device based on an association mode and a heterogeneous double graph, and a storage medium. The method comprises: obtaining a user information set, a point-of-interest information set, a point-of-interest category information set, and a user check-in time information set; constructing an association mode according to the user information set, the point-of-interest information set, the point-of-interest category information set, and the user check-in time information set; constructing a first heterogeneous graph and a second heterogeneous graph according to the association mode and the aforementioned sets; determining a first decomposition graph of the first heterogeneous graph and a second decomposition graph of the second heterogeneous graph based on vector embedding according to the association mode; determining a first information entropy and a first weight corresponding to the first heterogeneous graph and a second information entropy and a second weight corresponding to the second heterogeneous graph based on an information entropy weighting method; and determining a recommendation score of each point of interest according to the first information entropy, the first weight, the second information entropy, and the second weight. The application can improve the accuracy of point-of-interest recommendation.

The invention belongs to the technical field of electronic packaging material analysis, and discloses a modeling method for a polycrystalline porous microstructure of a sintered nano-silver material, and the method comprises the steps: firstly carrying out the binarization processing of an SEM image of sintered nano-silver, and separating out a pore region and a solid-phase region; then extracting the contour of the pore region to obtain a discrete point set on the contour; based on the discrete point set, constructing geometric constraint conditions representing pore boundaries; then, by taking the geometric constraint condition as a boundary which cannot be invaded, performing constrained Delaunay triangulation in the solid-phase region, and generating a constrained Delaunay triangulation network; and finally, extracting a geometric dual graph of the constrained Delaunay triangulation network, and generating a Voronoi graph. According to the method, the limitation of simplifying pore morphology or neglecting solid-phase anisotropy in a traditional method is overcome, and a real and reliable geometric model basis is provided for high-precision finite element simulation.

The invention provides a multi-scale multi-view clustering method and system based on deep learning, and the method comprises the steps: obtaining initial features of a plurality of visual angles, inputting the initial features of each visual angle into a preprocessing module, and enabling the preprocessing module to output and obtain a common structural feature matrix and a common node feature matrix; inputting the common structural feature matrix and the common node feature matrix corresponding to each view angle into a double-graph diffusion convolution module, and calculating the structural feature matrix and the node feature matrix corresponding to each view angle by the double-graph diffusion convolution module based on the common structural feature matrix and the common node feature matrix; obtaining an iteration structure feature matrix and an iteration node feature matrix corresponding to each view angle; the iterative structure feature matrix and the iterative node feature matrix are enhanced through a graph convolutional network of a high-order contrast learning module, and the iterative node feature matrix of each view angle is constructed into a common graph through a fine-grained multi-view fusion module.

This invention discloses a method for allocating resources in an Internet of Vehicles (IoV) network based on a primal-dual graph convolutional network. The method comprises the following steps: constructing a dynamic IoV network; converting the IoV network into a graph structure with node and edge features; establishing a mathematical model with the optimization objective of maximizing the weighted total packet reception rate of V2V links, characterizing the latency and reliability of V2V links using interruption probabilities, and converting latency and reliability constraints into power constraints using the series theorem, thereby combining the power allocation strategy with the latency and reliability constraints; constructing a graph neural network model to obtain a power allocation strategy that maximizes the packet reception rate of V2V links, and introducing a Lagrangian dual method to address quality of service requirements such as the packet reception rate. This invention uses a graph convolutional neural network as a power allocation strategy to maximize the weighted total packet reception rate of V2V links while ensuring quality of service requirements such as latency, reliability, and throughput of the V2V links.

The present invention provides a road network pattern recognition method and system based on unsupervised graph representation learning. First, the road network is modeled through a spatial dual graph, and the graph node features are designed based on cognitive inspiration. Then, the proposed graph neural network model is based on GAE, which can be trained in an unsupervised manner. At the same time, subgraph isomorphism count (SIC) is introduced in the road segment embedding learning stage and global context attention mechanism (GCA) is introduced in the graph embedding generation stage to enhance the model representation performance. Finally, the geometric similarity of graph-level embedding is used to identify road network patterns. Experimental results show that the present invention outperforms the classic road network recognition method in all indicators in the experimental area, and the classification accuracy is improved by more than 12%, which is even slightly better than the supervised baseline graph neural network method.

This invention proposes a path search method, electronic device, and storage medium that considers turning delays, belonging to the field of path search technology. It includes: S1. Establishing a mapping relationship based on the features of the original directed graph; S2. Establishing a mapping relationship from the original directed graph to the dual graph and creating the dual graph; S3. Adding edge weights of the dual graph based on the weights of the original directed graph; S4. Performing path search based on the dual graph; S5. Restoring the shortest path in the dual graph to a path in the original directed graph. This invention transforms the original directed graph into a dual graph, converting the edges of the original directed graph into nodes of the dual graph, and vice versa. The network size remains unchanged before and after the transformation, without altering the physical structure of the network. It also considers edge-to-edge turning delays, supporting path search starting with nodes and path search starting with directed edges. This solves the problem of low computational efficiency in existing path search methods that consider delays.

The invention discloses an inverse synthesis processing method and device, electronic equipment and a storage medium, relates to the technical field of inverse synthesis, and is used for improving the effect of inverse synthesis processing. The method comprises the following steps: acquiring an isomeric graph of a target molecule to be inversely synthesized and a dual graph corresponding to the isomeric graph; nodes in the heterogeneous graph represent atoms in the target molecule, and nodes in the dual graph correspond to a planar structure in the heterogeneous graph; determining feature vectors of a plurality of atoms in the target molecule according to the first node feature and the second node feature; the first node features comprise feature vectors of a plurality of first nodes in the heterogeneous graph; the second node features comprise feature vectors of a plurality of second nodes in the dual graph; and according to the feature vectors of the multiple atoms in the target molecule, performing inverse synthesis treatment on the target molecule to obtain a reactant molecule of the target molecule.

The application discloses a kind of based on bipartite graph's multi-port inverter topological derivation method, first, based on the mapping rule of bipartite graph, the component in multi-port inverter is mapped as the edge in bipartite graph, the circuit node in multi-port inverter, DC source, AC port are all mapped as the point in bipartite graph;Then based on the mapping rule of bipartite graph, basic circuit constraint is converted into graph constraint, so that the minimum degree of each point is obtained;Then, the degree of each point is determined in combination with handshake theorem, and then a bipartite graph satisfying graph constraint is derived;In addition, the topology type and more source of bipartite graph can be expanded, so that diversified effective bipartite graph is obtained;Finally, the diversified effective bipartite graph is converted into the topological graph of multi-port inverter.