46 results about "Constraint graph" patented technology

The application discloses a power metering drift correction method and system based on transformer area line loss and a medium, relates to the technical field of power meters, and comprises the following steps: constructing a four-level topology of a transformer area-branch-meter box-meter and attaching a communication quality label; calculating a measured line loss rate in layers, combining a physical constraint graph neural network to obtain a theoretical line loss; comparing a deviation and combining communication quality to determine suspicious nodes; performing time sequence decomposition on candidate meters to identify drift types; and adaptively verifying and grading correcting according to the drift types. The application solves the technical problems that it is difficult to determine the power metering drift node in the prior art, the drift type cannot be effectively distinguished, the operation and maintenance efficiency is low, and drift identification is prone to errors, and achieves the technical effects of improving line loss analysis accuracy by constructing a refined topology and a graph neural network, accurately identifying the drift type by combining a communication quality label and time sequence decomposition, and improving the reliability of power metering drift processing and the operation and maintenance efficiency.

The application discloses a power grid power flow prediction method and system based on a physical constraint graph convolution network, and relates to the technical field of power system power flow prediction.The method comprises the following steps: acquiring the topological structure and operation data of a power grid to be analyzed, constructing an adjacency matrix, and constructing a node feature matrix and an edge label matrix; inputting the node feature matrix into a graph convolution network for topological perception learning, and acquiring node representation; mapping the node representation into initial branch power flow prediction values through a bus-branch correlation mapping layer; adjusting the initial branch power flow prediction values through a node power balance correction layer by using a correction matrix, and obtaining corrected branch power flow prediction values; and combining a root mean square error and a physical regularization loss function to iteratively update model parameters, obtaining a power flow prediction model, and performing online reasoning.The application solves the problem of node power imbalance of a traditional data-driven model, and significantly improves the physical consistency and accuracy of power flow prediction.

The application discloses a source-load joint probability prediction method and system of a physically constrained graph attention network. The method collects multi-dimensional feature data of source-load nodes in a prediction area to construct an initial node feature matrix. A similarity matrix is generated through differentiable graph structure learning. A dynamic adjacency matrix is generated through normalization and introduction of a sparse mask. Spatial feature aggregation is performed through a multi-head graph attention network to obtain node spatial encoding. The node spatial-temporal hidden state is output through an encoder. The node spatial-temporal hidden state is input into a probability prediction head to output Gaussian distribution parameters of the source-load node power. A joint loss function is constructed. The joint loss function is used for soft constraint training to output a probability prediction result. Posterior projection hard constraint correction is performed in an inference stage to obtain a corrected prediction result. The application solves the problems of lack of physical consistency and inability to quantify uncertainty in the prior art.

The application relates to the technical field of data processing, and discloses a multi-source pipeline data registration and fusion labeling method for heterogeneous coordinate systems, which comprises the following steps: obtaining multi-source pipeline data and source coordinate system description information to establish a coordinate transformation model. Pipeline nodes, pipeline sections, connectivity relationships and semantic attributes are extracted, source and target pipeline network topology semantic constraint graphs are constructed, a registration error model is established, parameters are jointly solved through iterative optimization, and a globally consistent registration result is obtained. Pipeline nodes and pipeline sections from different pipeline data sources and corresponding in spatial position, topology structure and semantic attributes are clustered into unified pipeline objects and assigned with unified identification. A quality-driven attribute fusion model is constructed, geometric attributes are corrected within engineering constraints, classified attributes are weighted, and finally, a fusion labeling result is obtained. The application obtains unified pipeline objects which are globally consistent, topologically correct, semantically unified and provided with quality labeling.

This invention discloses an autonomous data processing system based on a deep mining framework, relating to the field of data processing technology. It includes: a semantic judgment module that extracts the target missing field, associated fields, and historical homologous field sequences from input records; determines the field constraint relationships between the target missing field and associated fields; and constructs a field constraint graph accordingly; calculates the missing position value, constraint break value, and historical offset value; and determines the missing semantic code based on these values; and writes the field constraint graph, the missing semantic code, and the missing position value, constraint break value, and historical offset value into a record-level state payload. This invention constructs a continuous data processing link around the target missing field, associated fields, and historical homologous field sequences; and establishes clear data flow and control flow relationships between vectors such as field constraint relationships, field constraint graphs, missing semantic codes, original data tracks, and completed data tracks.

This invention discloses a fixture generation method and system based on parametric templates and intelligent reasoning, relating to the field of fixture design technology. The method includes: acquiring a digital model of the workpiece and machining process documents; extracting workpiece geometric features, datum, process, and toolpath information; constructing a typified constraint graph, using nodes and edges to represent various tooling constraint relationships, and encoding the cutting load spectrum as dynamic constraint conditions for the constraint edges; retrieving fixture template skeletons from a parametric template library; solving the constraints on the template skeletons, calling a reduced-order physical simulation model in real time to complete mechanical evaluation, and relying on feedback signals to guide the search direction; after the solution converges, instantiating the template skeleton based on feasible solutions, and outputting the fixture model and solution certificate. This invention improves solution efficiency and solution reliability through simulation-embedded closed-loop optimization, relies on parametric templates to ensure design reusability, and the dual-element output enables full-process traceability of the design, meeting the high precision and compliance requirements of high-end manufacturing tooling design.

This application provides a method and related apparatus for dynamically constructing a surgical operation knowledge ontology based on surgical stage layering and a two-layer constraint graph neural network. The method includes: S1. Implementing a three-layer surgical stage layering: stage layer, operation layer, and action layer; S2. Extracting cross-layer operation relationships to obtain a set of cross-layer operation relationships; S3. Performing knowledge inference using a two-layer constraint graph neural network, with a hybrid score Score(r) = λ_rule·Score_rule(r) + λ_gnn·Score_gnn(r); S4. Performing ontology consistency verification using an OWL inference engine; S5. Incrementally expanding the surgical operation knowledge ontology; S6. Performing a three-dimensional quality score and outputting a knowledge ontology quality report. This application also provides related apparatus corresponding to the method, including devices, electronic devices, computer-readable storage media, and computer program products.

This invention discloses a method and system for constructing a SPEC parameter extraction structure vector model. The method includes the following steps: Step 1: Constructing a multi-constraint international standard knowledge graph; Step 2: Intelligently extracting SPEC document parameters based on the multi-constraint graph; Step 3: Constructing a multi-constraint structure vector model; Step 4: Constructing a sub-project experience database, and outputting the results of the multi-constraint structure vector model. This invention relates to the field of knowledge engineering interdisciplinary technology and can solve the problems of low parameter extraction efficiency, fragmented constraint relationships, poor generalization of vector models, and lack of sub-project experience systems in existing technologies.

This invention discloses a transfer learning method for motor temperature prediction based on a physically constrained graph neural network, comprising the following steps: Step S1: Constructing a labeled undirected graph based on the physical structure of the source motor; Step S2: Constructing a physically constrained graph neural network based on the labeled undirected graph; Step S3: Training the physically constrained graph neural network using the operating data and temperature data of the source motor to obtain a pre-trained model; Step S4: Constructing an undirected graph of the target motor based on its physical structure, and reusing the heat capacity parameters and thermal conductivity functions in the pre-trained model to obtain an initialized physically constrained graph neural network for the target motor; Step S5: Obtaining the temperature prediction model of the target motor. This invention, by embedding physical heat transfer laws into a transferable graph neural network structure, enables the rapid construction of a high-precision temperature prediction model with physical interpretability for new motors using only a small batch of data.

This invention discloses a method and system for optimizing administrative text element extraction based on natural language processing, belonging to the field of natural language processing technology. The method performs multi-granularity semantic structure parsing on administrative texts to construct multi-granularity semantic structure representation data capable of simultaneously representing word semantics, sentence functional attributes, and textual organization relationships; identifies semantic fragments of candidate administrative elements and constructs an administrative element structure constraint graph to depict the logical dependencies between administrative elements; introduces the structure constraint graph as explicit constraint conditions to perform structure-aware joint reasoning and consistency constraint optimization on administrative elements, suppressing element combinations that do not conform to administrative semantic logic, and outputting structurally consistent administrative text element extraction results; this invention can effectively improve the logical consistency, completeness, and business usability of administrative text element extraction results, and is applicable to application scenarios such as structured processing of administrative documents, government information management, and intelligent auxiliary decision-making.

The present application belongs to the technical field of nondestructive testing of concrete structures, and particularly relates to a method for identifying the topology of internal defects in concrete based on a graph neural network, comprising: obtaining three-dimensional detection data of a concrete test piece for voxel reconstruction and extracting a three-dimensional defect candidate; constructing an initial graph based on spatial adjacency relationships and extracting node features and edge features; recalibrating the adjacency relationships and edge weights based on the local graph structure topology index of the initial graph to obtain a topology-constrained graph, and performing message passing on the topology-constrained graph while updating the node embedding in combination with adaptive information aggregation weight; constructing a fitness function, optimizing the allocation matrix of nodes to clusters to determine graph coarsening mapping, and generating a multi-layer defect skeleton subgraph; extracting a cross-layer topology description vector from the multi-layer defect skeleton subgraph and outputting the classification and identification result of the defect topology structure. The present application can accurately classify and identify the topology of complex internal defects in concrete.

The application discloses an automatic parking and local obstacle avoidance trajectory generation method, comprising the following steps: identifying candidate parking spaces along a vehicle reference route, constructing a parking envelope, a neighborhood safety field and a reconnection point, forming a parking topological corridor and a parking space relationship constraint graph. A parking space relationship guide physical constraint bridge diffusion parking model PRB-PhysParkNet is constructed, a parking space relationship coding unit is used to extract parking space relationship and environmental state features, and a physical constraint bridge diffusion generation unit is used to generate candidate parking trajectories that meet vehicle kinematic constraints, boundary safety constraints, time-space collision avoidance constraints and parking final state constraints. The parking space relationship constraint graph, the parking topological corridor and the reconnection point are feedback corrected according to the physical violation quantity, the neighborhood time-space occupation quantity and the final state deviation, and re-planning is executed under the trigger of environmental state changes, and the target parking space, the target trajectory and the vehicle control quantity are output. Compared with the prior art, the safety, the executability and the environmental adaptability of the trajectory generation in a complex parking scene are improved.

The application provides a power distribution network risk identification and disposal method and system based on a power grid knowledge graph, which comprises the following steps: extracting an entity relationship triple set from a multi-source power grid data set, and constructing a time sequence event knowledge graph according to the entity relationship triple set; combining each structured knowledge graph information extracted from the time sequence event knowledge graph with a prompt word template to form model input data, and adopting a large language model to perform causal reasoning on the model input data to output a structured causal chain, thereby obtaining a causal chain set; converting the causal chain in the causal chain set into a causal directed graph, and converting the causal directed graph into a sequential constraint graph; searching for an optimal causal path corresponding to different types of power grid events in a search space defined by the sequential constraint graph to generate a causal model; acquiring real-time data flow of a power distribution network, and performing risk identification and risk root cause positioning on the real-time data flow based on the causal model; and generating an optimal disposal strategy according to the risk root cause.

The application relates to the technical field of artificial intelligence, in particular to a customer service reply method and system of a multi-modal fusion architecture and a storage medium, the method comprises the following steps: obtaining initial input data of a user, preprocessing the initial input data to generate a standard data source; performing decision compliance analysis on the standard data source to generate a decision judgment result; and generating a corresponding decision scheme according to the decision judgment result. Through a space-semantic double-constraint graph network, the application effectively solves the technical defects of fragmentation of multi-modal data recognition and low standardization degree of traditional intelligent customer service, overcomes the problem of insufficient adaptation of general recognition technology to government affairs formats, in the semantic constraint dimension, relying on the node association mechanism of the graph network, the spatial features are deeply fused with the government and enterprise professional semantic dictionary, the conversion of multi-modal data to structured policy elements is automatically completed, and the recognition accuracy is improved by more than 40% compared with traditional NLP technology.

This application discloses a multi-preference intelligent machining trajectory planning method based on semantic recognition using a large language model, relating to the field of machining trajectory planning technology. The method first constructs a CNC machine tool machining drawing dataset containing samples of machined workpiece drawings and corresponding intermediate representations of geometric contours. Based on this dataset, a large language model is trained to learn the mapping relationship between machining drawings and intermediate representations of geometric contours. The drawing of the workpiece to be machined is input into the trained model, generating corresponding intermediate representations of geometric contours end-to-end. A directed constraint graph of machining features is constructed based on the validated intermediate representations, encoding multiple machining preferences as computable parameters. The machining process decision model determines the machining implementation sequence, ultimately generating machining G-code adapted to the CNC system. This application enables intelligent parsing of unstructured semantics in engineering drawings, achieving adaptive process decision-making while considering multiple machining preferences, effectively improving the consistency of process planning and the level of machine tool automation.

The application discloses an AI technical consultation service system based on deep learning, which comprises a data acquisition and coding module, a structure constraint latent variable module, a structured decoding module, a chapter structure construction module, a chapter generation module, a consistency checking module and a regenerating control module.The data acquisition and coding module acquires multi-source data in the engineering consultation field and performs text coding, image block coding and table unit coding to form a multi-modal input sequence.The structure constraint latent variable module inputs the multi-modal input sequence into an improved Perceiver IO model and outputs a corrected latent variable representation.The structured decoding module generates a parameter matrix, a parameter relationship set and a chapter slot binding result.The chapter structure construction module constructs a chapter structure constraint graph and determines a chapter corresponding parameter set and a clause set.The chapter generation module generates chapter texts.The consistency checking module outputs a conflict chapter identifier.The regenerating control module outputs a consultation report text.The application improves the structural consistency and numerical accuracy of the technical consultation report.

This application relates to the field of resource scheduling, and provides a method, apparatus, electronic device, medium, and product for resource scheduling of edge nodes. The method includes: acquiring the node resource status of the edge nodes; acquiring task request needs; generating a resource heatmap based on the node resource status; parsing the task request needs for task constraints to obtain a task constraint graph; parsing the task request needs for task priorities to obtain a task priority constraint graph; and scheduling resources for the edge nodes based on the resource heatmap, task priority constraint graph, and task constraint graph to obtain a resource scheduling result. The edge node resource scheduling method provided in this application can improve resource utilization by real-time acquisition of the node resource status of edge nodes, generating a resource heatmap based on the node resource status, and determining the resource scheduling result of the edge nodes for task request needs based on the resource heatmap, combined with the task constraints and task priorities of the task request.

The application provides a hyperactivity disorder auxiliary diagnosis method based on a multi-flow graph neural network and space-time constraints, comprising: acquiring motion sensor data of a user in a human-computer interaction scene, and extracting time domain statistical features of the motion sensor data; acquiring a space-time constraint graph of the human-computer interaction scene based on the time domain statistical features; acquiring integral graph vector representations of the space-time constraint graph; and fusing all the integral graph vector representations to obtain a classification result of the human-computer interaction behavior of the user. The application also provides a hyperactivity disorder auxiliary diagnosis system based on a multi-flow graph neural network and space-time constraints, and a data processing system for realizing hyperactivity disorder auxiliary diagnosis based on a multi-flow graph neural network.

This invention discloses an agent intent recognition method based on semantic understanding and contextual reasoning, comprising the following steps: receiving user input information and contextual data related to the user input information, and performing standardization processing; performing semantic parsing on the standardized input data to generate a candidate intent set and extracting a historical intent set; constructing a contextual constraint graph associated with the candidate intent set based on the standardized input data, and constructing an intent evolution graph; inputting the candidate intent set, historical intent set, contextual constraint graph, intent evolution graph, and standardized business rule information into an intent evolution CRmod model to obtain the final intent recognition result; determining whether the final intent recognition result meets the clarification conditions, generating clarification request information and feeding it back to the user if the conditions are met, and triggering corresponding task execution or business processing if the conditions are not met. This invention employs dual-graph collaborative reasoning to achieve agent intent recognition.

This application belongs to the interdisciplinary field of computer application technology and aerospace equipment technology, and specifically relates to an intelligent technical document compilation system and method based on evidence constraints. The system includes a data governance module, a compilation constraint graph construction module, an automated generation module, an intelligent review module, an intelligent retrieval module, a change impact analysis module, and an interface integration module. The system converts multi-source equipment technical data into source record objects, chapter unit objects, and evidence objects, constructs a chapter constraint graph, and generates chapter task packages containing template skeletons, mandatory entities, mandatory parameters, and evidence sets. Based on this, it outputs initial drafts of the document with evidence identifiers, and forms candidate release versions through rule review, graph consistency review, and semantic review. When the technical status changes, it identifies affected chapters and recompiles them accordingly. This invention improves the traceability, consistency, and timeliness of technical document compilation, review, retrieval, and updating.

This invention discloses an intelligent power distribution system and method, comprising: collecting and processing multi-source data to construct an input sample set, segmenting loads and power sources, and generating a mapping matrix; processing the sample set using a physical constraint graph neural ODE model, introducing the mapping matrix into the input layer, and outputting extended prediction results; inputting the extended prediction results into a modulation module, using the mapping matrix to perform cross-phase scheduling and generate a scheduling scheme; implementing a dual-time-domain event-triggered mechanism control scheme, with slow-loop rolling optimization and fast-loop rapid adjustment to generate control commands; issuing the scheduling scheme and control commands, with the power distribution controller, energy storage device, and phase modulator executing the operations; collecting execution feedback, updating the ODE model, mapping matrix, and event thresholds to form an adaptive closed loop. This invention achieves intelligent, controllable, and efficient operation of a distributed power distribution system through segmented phase energy modulation, physical constraint graph neural ODE modeling, and a dual-time-domain event-triggered mechanism.

This invention proposes a multi-step fault prediction method for chemical processes based on a knowledge-enhanced graph Transformer. The steps are as follows: constructing a multivariate time series and processing it with a sliding time window to obtain input samples; constructing a mechanism constraint graph and assigning weights to node pairs to form a prior adjacency matrix; constructing a learnable adjacency matrix based on the input samples and fusing it with the prior adjacency matrix; inputting the node feature matrix into a graph convolutional network to obtain a spatial feature matrix; adding a position encoding vector to the input samples and inputting them into the Transformer module to obtain the temporal dependency feature matrix for each time step; concatenating the inputs into a gating network and performing weighted fusion to obtain a fused feature matrix; using the obtained fused feature matrix to generate the next prediction, and using the prediction results and historical sequences for inference to achieve multi-step prediction. This invention significantly improves the early fault identification and warning capabilities of industrial processes under complex operating conditions while maintaining model interpretability and operational stability.

The application discloses a kind of power single-line diagram equipment marking method and device, belong to marking field, and the present application is by obtaining work ticket text data and power single-line diagram file, utilize the work ticket analysis model constructed based on historical work ticket data and preset power equipment hierarchical subordinate rule, work ticket text data is converted into the hierarchical path object containing power equipment entity and its hierarchical subordinate relationship.Subsequently, power single-line diagram file is parsed, and the drawing data model containing electrical equipment node, equipment interconnection and equipment space geometric information is constructed.Then, based on hierarchical path object, hierarchical constraint graph traversal is executed to drawing data model, and candidate equipment set is screened out, and target equipment is determined by spatial proximity check.Finally, according to work ticket text data, target equipment is marked, and the present application effectively solves the problem that power system cannot be accurately and efficiently marked in prior art.

The application provides an IP intelligent derivation method and system for clothing and household articles, relates to the technical field of intelligent design, and comprises the following steps: IP representation is subjected to manifold transformation and decomposition by constructing a cross-modal mapping space, guided projection is carried out in combination with function and structure constraints to obtain adaptive representation, and candidate forms are generated and screened. Manufacturability evaluation and conflict solving are carried out based on a manufacturing constraint graph, and a physical implementation scheme is obtained. The application realizes intelligent form derivation and manufacturability guarantee from IP to product.

The application provides a time-dependent label constraint efficient query method based on partition tree decomposition, relates to the technical field of data processing, and is used for path planning application in an intelligent transportation system and comprises the following steps: acquiring road network information and generating a road data graph; performing reduction on the road data graph based on a label partition tree decomposition algorithm and generating a tree structure; constructing a parameter separation index mechanism and converting the tree structure into a parameter separation index; acquiring a departure point and a destination point and taking the departure point and the destination point as a source vertex and a target vertex respectively; and calculating a shortest path from the given source vertex to the target vertex based on the parameter separation index. Through the design of the label partition tree decomposition and the parameter separation index, the application greatly reduces the index storage redundancy of a time-dependent label constraint graph, can efficiently and compatibly support ordered label constraints and time-dependent weights to adapt to an intelligent transportation scenario, and can linearly expand with the scale of a data set and maintain stable performance in a network with more than ten million vertices.

This invention relates to the field of smart distribution network and power equipment condition monitoring technology, and particularly to a fault early warning method, system, and ring network box for integrated primary and secondary ring network boxes. The method constructs a dynamic physical consistency constraint graph based on electrical topology and switch states, verifies multi-source time-series data, extracts conflicting data when constraints are violated, and updates it attenuation to obtain a basic confidence level. Simultaneously, it combines the real-time computing power of edge computing units to assess the expected inference time. Using the basic confidence level and expected inference time as joint gating conditions, when the confidence level exceeds the limit or the inference faces timeout risk, a second early warning path is triggered to output a conclusion within the deadline and generate early warning traceability data. This invention couples data consistency with edge computing power risk, avoiding decision delays and distortions, and ensuring the baseline reliability and traceability of early warnings under complex operating conditions.

The application discloses a kind of unmanned aerial vehicle aerial image road detection methods based on deep learning, it is related to deep learning technical field, including, the road multisource perception feature of extracting road block parameter set is modulated to road multisource perception feature by gate feature, and topological relationship feature map is generated;Topological relationship feature map is input dynamic perception model, global context modeling is carried out in road topological perception layer, time phase coupling is carried out in geometric constraint segmentation layer, and road probability graph and geometric constraint graph are generated;Topological relationship reasoning is generated to road probability graph and geometric constraint graph with road optimization candidate set, and road optimization candidate set is structured coding, and road structured data set is output.The dynamic perception model and gate feature modulation of the present application through deep learning improve the dynamic adaptability and time consistency of road detection and image detection.

The application discloses a kind of multimodal data fusion analysis and processing method and system, it is related to edge computing technical field;The method includes obtaining the control state sequence of target equipment, current response sequence and image sequence, extracts each modal candidate response time, constructs edge disturbance description quantity and candidate event fragment;Call event template to build cross-modal propagation sequence constraint graph, inverse solve virtual time scale sequence and form sequence deviation quantity;According to virtual time scale sequence, reconstruct event level synchronization fragment, extract logic jump feature, current dynamic feature and visual motion feature, generate event attribution credibility and event fusion representation quantity;According to event attribution credibility, output fusion analysis result, and feedback type update is carried out to propagation time delay interval, the application can improve the accuracy, stability and self-adaptive updating capability of multimodal event analysis.

This invention provides a method for detecting conflicts in the construction layout of microgrid power-grid-load-storage equipment based on 3D vision and graph optimization algorithms, belonging to the interdisciplinary field of swarm intelligence and computer vision. The method involves collecting 3D point clouds from the construction site and preprocessing them to obtain registered point clouds, simultaneously extracting equipment and structural component data, and retrieving electrical safety distance and maintenance access specification data. A 3D semantic model is constructed and bounding boxes are generated. A physical constraint graph is built using the bounding boxes as nodes and safety distances as edge weights, and input into an improved graph convolutional network to complete layout conflict detection. Access widths are obtained, and maintenance accessibility evaluation indicators are calculated based on specifications. A spatial constraint matrix is ​​established based on safety distances, with maintenance accessibility as a hard constraint. The goal is to reduce layout conflicts and improve space utilization. Heuristic search iterative optimization is used to output the optimal equipment layout scheme. This solution can provide intelligent decision support for the construction layout of microgrid power-grid-load-storage equipment.