28 results about "Composite graph" patented technology

A method for training a prediction model for a lower urinary tract symptom risk score is disclosed. A method for training a prediction model for a lower urinary tract symptom risk score according to an embodiment comprises the steps of: obtaining numerical data of a urinary system; obtaining graph data of the urinary system; obtaining composite graph data in which the numerical data is reflected in the graph data; obtaining a numerical feature map corresponding to the numerical data; obtaining a composite graph feature map corresponding to the composite graph data; and training the prediction model for the lower urinary tract symptom risk score on the basis of the numerical feature map and the composite graph feature map.

This invention provides a method and system for constructing a reliable evaluation dataset for power vision models, belonging to the field of intelligent recognition technology for power vision models. The construction method includes: acquiring collected on-site images and obstacle images; constructing virtual 3D models of power equipment and virtual obstacle models; marking the virtual 3D models and virtual obstacle models to obtain corresponding labels; acquiring multiple images from a virtual perspective on the composite model of the virtual 3D models and virtual obstacle models; determining the positional relationship between the virtual 3D models and virtual obstacles in the images based on the images; and generating a composite image and corresponding composite labels based on the positional relationship, the on-site images, and the obstacle images to form a dataset.

The invention provides a preparation method of a composite patterned substrate, the composite patterned substrate and an LED chip, and the method comprises the steps: forming a plurality of holes in the surface of a sapphire substrate, and carrying out the gluing, exposure and development processing, and obtaining a preset pattern; depositing a SiOF layer on the substrate; removing the preset pattern below the SiOF layer through heat treatment to form a cavity; transferring a pattern on the substrate through a nanoimprint technology; and performing inductively coupled plasma etching to form a composite pattern structure. A cavity structure is formed in a sapphire substrate through laser drilling and heat treatment, the light extraction efficiency of an LED is improved by combining with a composite pattern constructed by a nanoimprint technology and utilizing a multi-interface light scattering effect, the propagation direction of light rays can be guided by combining a patterned structure with a cavity, light emitting angle distribution is optimized, and light emitting is more uniform or the directivity is more concentrated; epitaxial growth stress is released through the cavity, dislocation density is reduced, epitaxial layer cracks are effectively inhibited, and crystal integrity is improved.

This invention provides a method, system, device, and storage medium for smart contract vulnerability detection, belonging to the fields of blockchain security and artificial intelligence technology. It includes extracting an abstract syntax tree, control flow graph, and data dependency graph from the original smart contract code; obtaining a composite graph object based on the abstract syntax tree, control flow graph, and data dependency graph; processing the composite graph object based on node semantics and function call relationships to obtain a hierarchical heterogeneous graph; constructing a heterogeneous graph Transformer network; using a heterogeneous graph Transformer encoder to process the hierarchical heterogeneous graph for encoding node and edge information, obtaining node representation vectors that fuse semantic and structural information; extracting multi-scale features from the node representation vectors; fusing the extracted multi-scale features using learnable weights; and processing the fused multi-scale features using a base classifier cluster to obtain vulnerability type prediction results and confidence levels. This invention achieves full-process optimization from structural modeling and semantic fusion to decision reasoning, significantly improving the accuracy and interpretability of smart contract vulnerability detection.

This invention provides a fully printed flexible photoresponse array and a parallel pattern recognition method, belonging to the field of semiconductor optoelectronic device technology. Firstly, this invention provides a fully printed flexible photoresponse array, including a flexible substrate and multiple differential dual-branch photoresponse pixels disposed on the flexible substrate. Each differential dual-branch photoresponse pixel includes: a common source electrode, a first branch drain electrode, a second branch drain electrode, a first composite channel branch, a second composite channel branch, and a light-shielding layer disposed above the second composite channel branch. Based on a special structural design and connection relationship, this invention solves the problems of insufficient stability in flexible array recognition, significant background light interference, and insufficient utilization of temporal information in existing technologies. Simultaneously, based on the above structure, a parallel pattern recognition method is provided to achieve parallel recognition and memory retention representation of multiple characters, multiple graphic units, or composite patterns.

This invention discloses a data intelligent sensing and optimization scheduling system and method for aluminum processing, belonging to the field of aluminum processing optimization scheduling technology. The method specifically includes: collecting multi-source heterogeneous data on aluminum processing, preprocessing the collected multi-source heterogeneous data, establishing a composite graph, dynamically generating a feature matrix using the correlation between different types of data, constructing a scheduling model for aluminum processing, performing pre-simulation using a virtual production twin, optimally configuring production resources for aluminum processing, applying the optimization scheduling model to aluminum processing in real time, and continuously optimizing the scheduling model using a reinforcement learning strategy for dynamic adjustment. This application can automatically optimize the allocation and scheduling order of resources according to changes in the manufacturing environment, improving production efficiency and reducing energy consumption and waste.

The invention relates to an antenna array synthesis directional diagram estimation method based on a multi-unit antenna actual measurement directional diagram and application, and the method comprises the following steps: in a standard test environment, measuring an azimuth plane or pitching plane directional diagram of each antenna unit, recording angle gain or power data, and obtaining actual measurement data; establishing a coordinate system, determining antenna reference positions and inter-unit parameters, dividing every two antennas into one group, and scoring groups and parameters; according to the directivity multiplication principle, a double-antenna composite graph computational logic (estimation formula) is established, and data simulation is substituted to obtain a composite graph; and taking the composite diagram as a quasi-unit diagram, and repeating the operation to complete the estimation of the composite directional diagram of the multi-unit antenna array. According to the method, the defects of complex modeling, large calculation amount and poor adaptability of a traditional theory multi-unit antenna array can be avoided, the blank in the prior art is filled, the estimation efficiency and precision can be effectively improved, the requirement for rapid and accurate analysis of the performance of a multi-antenna system in engineering is met, and reliable support is provided for design optimization of the antenna array.

The present invention proposes a vulnerability detection method based on a dual attention mechanism and code similarity, which belongs to the field of vulnerability detection technology. It includes: extracting the program dependency graph corresponding to the unknown function to be tested, expanding the abstract syntax tree corresponding to the statement to each node of the program dependency graph to form a composite graph; using the marked vulnerability statement as a slicing criterion, finding the slicing node in the program dependency graph as the root node, expanding the abstract syntax tree to form a slice composite graph; learning the syntax and semantic information within the statement, generating an initial embedding vector of the statement node composed of semantic and syntax feature vectors; learning the structural features of the source code, obtaining the slice composite graph and the vector representation of the composite graph, and calculating the similarity between the two; when the similarity exceeds a certain threshold, the target function is defined as a potential vulnerability function. This method solves the problem of vulnerability detection methods that lack vulnerability information that can effectively exploit known vulnerability codes while ignoring or even eliminating redundant irrelevant information in the code.

The application provides a drug-target protein binding affinity prediction method and system, relates to the technical field of biological information technology, and comprises the following steps: learning local information of an atomic composite graph at a local level by using an isometric graph neural network; fully utilizing three-dimensional space structure information by using the isometric graph neural network, and keeping rotation and translation of coordinates unchanged during message transmission; then using a node feature extraction model and an atomic interaction matrix to learn deep-level interaction of the atomic composite graph at a global and long-distance level, improving interpretability of feature learning of the protein and the ligand, and making prediction performance of the drug-target protein binding affinity prediction model optimal. The application combines interaction of each distance level of the drug-target protein by fitting a dynamic process of drug-protein combination, simultaneously considers rotation and translation invariance of atomic mutual combination, and effectively improves accuracy of drug-target protein binding affinity prediction.

The application discloses a network security-oriented constraint-driven fast topology presentation method, device and medium, which comprises the following steps: connecting subgraphs of a composite graph to form an integral connected graph, generating a preliminary layout based on the connected graph; performing rotation and reflection transformation operations on the entire layout to improve the compatibility of the entire layout with constraints, and executing various constraint conditions to obtain a preliminary constrained layout; on the basis of the preliminary layout, eliminating the overlap between nodes and the nesting relationship in the graph, and maintaining various constraint conditions executed previously to refine and optimize the entire layout; setting a low cooling factor, gradually eliminating the overlap, maintaining the nesting, and maintaining the constraints, and finally obtaining a high-quality layout; and presenting the final network topology on a canvas. The application is suitable for graphs that need to display complex node relationships and hierarchical structures, and the algorithm can ensure high-quality layout while improving the layout speed.

Embodiments of the present specification disclose an electric vehicle fast charging method and device, electronic equipment and storage medium. The method comprises: after receiving a charging instruction, obtaining historical test data; constructing a current-capacity hyperbolic curve composite graph and a current-temperature hyperbolic curve composite graph in turn through a first test data set and a second test data set; determining a mutation intersection capacity value and a mutation intersection temperature value respectively; continuously monitoring the charging state of the target battery; comparing each real-time charging parameter with the mutation intersection capacity value and the mutation intersection temperature value; and finally determining the charging mode of the target battery in the current charging stage based on the comparison result, switching between interval charging mode and linear charging mode to keep the target battery at a high battery charging current throughout the charging process, reducing the charging time of the target battery, achieving the purpose of electric vehicle fast charging, and improving user satisfaction.

We provide the system. [Solution] A means for the user to select building elements, A means for the server to search the database based on the selected element and retrieve the corresponding image data, A means of displaying image data acquired by the terminal to the user, A means of reflecting user-selected elements onto the drawing, A means for generating a composite drawing based on a drawing reflected by a server, A means by which the terminal displays a composite diagram, A system that includes this.

This invention provides a fully printed flexible photoresponse array and a parallel pattern recognition method, belonging to the field of semiconductor optoelectronic device technology. Firstly, this invention provides a fully printed flexible photoresponse array, including a flexible substrate and multiple differential dual-branch photoresponse pixels disposed on the flexible substrate. Each differential dual-branch photoresponse pixel includes: a common source electrode, a first branch drain electrode, a second branch drain electrode, a first composite channel branch, a second composite channel branch, and a light-shielding layer disposed above the second composite channel branch. Based on a special structural design and connection relationship, this invention solves the problems of insufficient stability in flexible array recognition, large background light interference, and insufficient utilization of temporal information in existing technologies. Simultaneously, based on the above structure, a parallel pattern recognition method is provided to achieve parallel recognition and memory retention representation of multiple characters, multiple graphic units, or composite patterns.

The invention relates to the field of image processing, and discloses a power grid wiring diagram composite primitive construction method, system and device and a medium. The method comprises the steps of obtaining a power grid wiring diagram, generating a primitive connecting line segmentation mask image through a primitive connecting line segmentation model, and generating a plurality of basic primitive rectangular frames and corresponding basic primitive types through a basic primitive detection model; segmenting the mask image based on each basic primitive rectangular frame, the corresponding basic primitive type and the primitive connecting line, and constructing a refined composite primitive refining frame and the corresponding composite primitive type by using a composite primitive construction algorithm, so as to obtain a power grid wiring diagram recognition result; and recording and storing an identification result list according to the identification result of the power grid wiring diagram. According to the method, the problems of low detection rate and high false detection rate of the composite primitive in the prior art are effectively eliminated, the recognition precision and speed of the composite primitive are remarkably improved, and the practical value is relatively high.

A non-transitory computer readable medium includes stored instructions, which when executed by a processor, cause the processor to acquire a plurality of graphs that model results of a plurality of bug hunting searches for a register transfer level design, wherein the plurality of bug hunting searches determines a design error in the register transfer level design, construct a composite graph using the plurality of graphs, wherein the composite graph models a plurality of causal relationships between a plurality of detected bugs and a plurality of helper properties covered by the plurality of bug hunting searches, and execute an additional bug hunting search whose parameters are configured based on the composite graph.

Threat report processing is provided. A composite graph for a threat report corresponding to a cybersecurity threat is generated by merging subgraphs corresponding to cyberattack-related sentences identified in the threat report into one or more graphs. The composite graph of the threat report corresponding to the cybersecurity threat is analyzed to identify indicators of compromise associated with system entities and indicators of behavior associated with the subgraphs comprising the composite graph. Descriptions of the indicators of compromise associated with the system entities and the indicators of behavior associated with the subgraphs comprising the composite graph are generated based on analyzing the composite graph of the threat report corresponding to the cybersecurity threat. A threat hunting query corresponding to the cybersecurity threat is generated based on the descriptions of the indicators of compromise associated with the system entities and the indicators of behavior associated with the subgraphs comprising the composite graph.

The application provides an industrial time series anomaly detection method, device and computer equipment, and belongs to the technical field of industrial time series anomaly detection. The method comprises the following steps: acquiring industrial time series data; converting the time series data into graph structure data as model input data to input a reconstruction model, to obtain reconstruction data, wherein the reconstruction model comprises a number knowledge composite graph attention network; if the difference degree between the reconstruction variable data of a certain timestamp in the reconstruction data and the original variable data corresponding to the timestamp in the model input data reaches a preset anomaly threshold, it is determined that the original variable data corresponding to the timestamp is in an abnormal state. According to the knowledge graph related to the industrial system, the reconstruction model extracts a superior representation of the variable interaction network based on knowledge, and fuses it with the sequence-based time series dependence, thereby reducing the risk of learning pseudo-correlation by the model, and improving the anomaly detection performance and anomaly explainability of the industrial time series.

The application relates to the field of image processing and discloses a power grid wiring diagram composite graph element construction method, system, device and medium. The method comprises the following steps: acquiring a power grid wiring diagram, generating a graph element connection segmentation mask image through a graph element connection segmentation model, and generating a plurality of basic graph element rectangular frames and corresponding basic graph element types through a basic graph element detection model; based on the basic graph element rectangular frames, the corresponding basic graph element types and the graph element connection segmentation mask image, and by using a composite graph element construction algorithm, a refined composite graph element refined frame and corresponding composite graph element types are constructed, and then a power grid wiring diagram recognition result is obtained; and a recognition result list is recorded according to the power grid wiring diagram recognition result, and the recognition result list is saved. The application effectively eliminates the problems of low composite graph element detection rate and high false detection rate in the prior art, significantly improves the recognition accuracy and speed of the composite graph element, and has strong practical value.

A non-transitory computer readable medium includes stored instructions, which when executed by a processor, cause the processor to acquire a plurality of graphs that model results of a plurality of bug hunting searches for a register transfer level design, wherein the plurality of bug hunting searches determines a design error in the register transfer level design, construct a composite graph using the plurality of graphs, wherein the composite graph models a plurality of causal relationships between a plurality of detected bugs and a plurality of helper properties covered by the plurality of bug hunting searches, and execute an additional bug hunting search whose parameters are configured based on the composite graph.

The invention provides a combined image checking method and system for a sorting machine, a computer and a storage medium. The combined image checking method for the sorting machine comprises the steps that a wafer is scanned, a combined image file is formed, a plurality of preset verification coordinates are set, and whether coordinate data of the combined image file contain all the preset verification coordinates or not is judged; if the coordinate data of the composite image file does not contain all the preset verification coordinates, determining a plurality of preset verification coordinates contained in the coordinate data as a plurality of to-be-verified coordinates; identifying the crystal grains according to the to-be-verified coordinates, and judging whether the to-be-verified coordinates meet a first verification standard or not; if the to-be-verified coordinate accords with the first verification standard, identifying a crystal grain according to an auxiliary verification coordinate, and judging whether the to-be-verified coordinate accords with a second verification standard; and calculating a non-wafer verification success rate based on the number of the successfully verified coordinates and the number of the to-be-verified coordinates. According to the method, various wafer conditions can be effectively compatible, misjudgment of image combination inspection is reduced, and manual intervention is greatly reduced.

The invention discloses a data intelligent sensing and optimal scheduling system and method for aluminum processing, and belongs to the technical field of aluminum processing optimal scheduling, and the method specifically comprises the steps: collecting multi-source heterogeneous data of aluminum processing, carrying out the preprocessing of the collected multi-source heterogeneous data, building a composite atlas, and carrying out the analysis of the composite atlas; the method comprises the following steps of: dynamically generating a characteristic matrix by utilizing an incidence relation among different types of data; constructing a scheduling model of aluminum processing; performing rehearsal by utilizing a virtual production twinborn body; performing optimal configuration on production resources of aluminum processing, applying an optimal scheduling model to aluminum processing in real time, continuously optimizing the scheduling model by utilizing a reinforcement learning strategy, and performing dynamic adjustment. The resource allocation and scheduling sequence can be automatically optimized according to the change of the manufacturing environment, the production efficiency is improved, and energy consumption and waste are reduced.

The invention discloses a coal mine multi-disaster risk area intelligent division method based on a graph neural network, and the method comprises the following steps: collecting multi-disaster monitoring data and space structure information, and forming original monitoring data; constructing a composite graph structure based on the original monitoring data; according to the composite graph structure, directional input is obtained through directional association; based on the directional input, a Hodge-Laplacian directional information transmission mechanism is introduced, and an edge state updating result is obtained; aggregating the edge state updating results to the corresponding connection nodes, and outputting node states; based on the node state, performing fusion to generate enhanced node representation; constructing an anisotropic thermonuclear diffusion diagram model, and generating a multi-disaster risk propagation diagram; and according to the multi-disaster risk propagation graph, performing normalization and threshold division according to node scores, and outputting a coal mine multi-disaster risk area division graph layer with a continuous boundary. According to the invention, intelligent division of multiple disaster risk areas of the coal mine is realized.

The invention relates to the technical field of big data processing, in particular to a hardware fault intelligent inspection system based on a BMC log, and the system comprises a topology event analysis module, a degradation state mapping module, a macroscopic state inference module and a microscopic state deepening module. According to the method, the number of events and the highest severity level in the log stream are spliced in real time, a dynamic component event vector is constructed, a composite graph model is constructed by using a hardware topological structure and a hierarchical state space, an isolated fixed threshold is not depended on any more, and the macro operation state posterior probability of the component is deduced on the graph model through a probability model; in addition, a corresponding classifier can be automatically selected according to the macroscopic state, the microscopic degradation process of the analysis component is deepened, and the recognition precision of early faults and complex relevance anomalies is remarkably improved.

A method of analyzing film on a substrate comprises receiving surface profile data obtained from measurements of a substrate having a plurality of discrete regions with one or more film layers; extracting, based on a predetermined pattern of the discrete regions, a plurality of parameters from the received surface profile data, the plurality of parameters comprising one or more parameters of a film layer of each discrete regions, and displaying a user interface. The user interface may comprise a plurality of individual graphs each illustrating the one or more parameters of the one or more film layers for a corresponding subset of the plurality of discrete regions, and a composite graph illustrating the one or more parameters of the one or more film layers for each discrete region of the plurality of discrete regions, wherein the composite graph corresponds to the plurality of individual graphs being overlaid together.

This invention discloses a method and system for risk warning and traceability of aflatoxin in corn feed, relating to the field of feed safety technology. The risk warning and traceability method includes the following steps: real-time collection of storage environment data, combined with supply chain source quality information, to construct a dual composite graph structure of physical adjacency and biological propagation; employing a source quality-driven dynamic gated spatiotemporal graph neural network model to achieve 5-7 day advance risk prediction; quantifying the proportion of responsibility for source quality, spatial propagation, and self-management based on a rate-of-change sensitive spatiotemporal integral gradient attribution method; and outputting a cost-optimized precise intervention plan through digital twin simulation and reinforcement learning optimization. This invention solves the technical problems of fragmented mechanistic data, neglect of spatial topology, and insufficient traceability interpretation in existing technologies, significantly improving the accuracy, timeliness, and interpretability of aflatoxin risk warning.

This invention discloses a method for dynamic modeling of bidirectional traffic flow based on a deep spatiotemporal graph neural network, belonging to the field of intelligent transportation and data mining. The method includes the following steps: S1, constructing a dynamic traffic flow graph to capture the local spatiotemporal dependencies of the traffic flow; S2, after obtaining the local spatiotemporal dependencies, using a temporal convolutional layer to extract temporal features; S3, after obtaining the temporal features, using a selection gate to filter the effective information obtained in S2 to obtain the final output; S4, after obtaining the result in S3, using a composite graph convolutional module to extract spatial features; S5, fusing the bidirectional traffic flow and obtaining the final prediction result through a prediction layer; S6, calculating the model loss to determine whether the model meets the requirements. This invention addresses the problems of insufficient bidirectional traffic flow modeling, inadequate capture of dynamic dependencies, and poor adaptability to complex traffic scenarios faced by existing technologies in traffic flow prediction, thereby improving the accuracy and applicability of traffic flow prediction.