189 results about "Topology construction" patented technology

The invention relates to the technical field of construction stage management, in particular to a BIM-based road construction digital management system, which comprises a component parameter acquisition module, a process topology construction module, a state trigger mapping module, a logic conflict judgment module and a scheduling path adjustment module. According to the method, through continuous collection and analysis of the component state, dynamic monitoring of the connection state, geometric intersection and contact time is achieved, the construction change sensing capacity is improved, the dependency relationship is constructed in combination with component space adjacency and time overlapping, the dispatching path and field matching degree is enhanced, and through linkage updating of the component state and model attributes, the construction efficiency is improved. According to the method, the execution state is visual and controllable, scheduling reconstruction is carried out after time dependence conflicts are recognized, synchronous updating of model views is completed, a closed-loop dynamic scheduling mechanism is formed, the task allocation accuracy, the flow connection efficiency and the field scheduling response capability are integrally improved, and the real-time performance and the intelligent level of construction management are enhanced.

The invention relates to the technical field of new energy storage regulation and control, and discloses an electricity-hydrogen coupling energy storage system and method for a new energy field station. The system comprises a feature analysis unit, a topology construction unit, a regulation and control domain positioning unit and a decision optimization unit. The feature analysis unit receives operation condition information of the new energy station, and generates a condition feature vector according to real-time load fluctuation, a historical output curve and an equipment operation database by using a deep feature extraction network; the topology construction unit constructs an energy topological graph for distributed energy storage equipment and hydrogen production equipment in the electricity-hydrogen coupling system, dynamically updates a graph structure and a node state based on a real-time monitoring mechanism, and deploys the graph structure and the node state in a heterogeneous computing framework; a regulation and control domain positioning unit positions a matched regulation and control domain in a dynamic representation learning mode according to the working condition feature vector; and the decision optimization unit utilizes a multi-source decision fusion algorithm to screen regulation and control instructions in a regulation and control domain and generate a scheduling sequence, so that the adaptability and the operation efficiency of the system to working conditions are improved.

The invention relates to an underground pipe network topological relation analysis method and device, and belongs to the technical field of underground pipeline management.The underground pipe network topological relation analysis method comprises the steps that a spatial index structure is constructed, an adjacent node set of each node is determined based on the spatial index structure, and a spatial ray corresponding to each node is constructed; determining a correlation score between each node and the nodes in the corresponding adjacent node set based on the space ray and pipe orifice multi-dimensional information corresponding to each node; determining a candidate connection node of each node based on a correlation score and a score threshold value between each node and the nodes in the corresponding adjacent node set, and constructing an initial pipe network topology based on the candidate connection node of each node; and optimizing the initial pipe network topology based on a preset topology construction logic to obtain the pipe network topology. According to the method, the topological relation analysis efficiency of the underground pipe network is improved, and meanwhile, the topological relation analysis accuracy is guaranteed.

The invention relates to the technical field of intelligent operation and maintenance, in particular to a fault insight and self-healing method and system based on chain control. According to the chain control-based fault insight and self-healing method, indexes, logs and tracking data of a micro-service environment are dynamically collected and fused; the method comprises the following steps: dynamically constructing and visualizing a service dependent topology, realizing dynamic topology construction and intelligent visual rendering, performing intelligent root cause analysis and predictive diagnosis based on multi-dimensional data, realizing self-adaptive early warning, generating an intelligent decision, performing closed-loop learning optimization based on an execution result, and realizing automatic fault repair and continuous optimization. According to the fault insight and self-healing method and system based on chain control, a predictive early warning mechanism is established, automatic fault self-healing is realized, and through intelligent and automatic means, manual intervention is reduced, the operation and maintenance cost is reduced, the average repair time is shortened, the service continuity is guaranteed, and the system availability is improved.

The invention discloses a method and equipment for determining a wiring state and a storage medium. The method comprises the following steps: receiving target image data containing image information at a wiring position, performing target detection and feature extraction processing on the target image data, automatically identifying a connecting element (a first target) and an identifier (a second target) thereof in an image, and simultaneously extracting an appearance visual feature (a first feature) of the connecting element and an identification code feature (a second feature) of the identifier; and performing topology construction processing based on the obtained target and characteristics to generate topological graph data capable of accurately representing the connection relationship of the electrical devices. And finally, comparing the topological graph data with a preset standard wiring template topological graph, and automatically generating judgment information for indicating whether the wiring connection state is correct or not according to a topological matching result. Therefore, misjudgment caused by artificial fatigue, negligence and experience difference is effectively avoided.

The invention relates to the technical field of supply chain monitoring, and particularly discloses an automatic sensor detection method and system based on a supply chain, and the method comprises the steps: generating a hierarchical pressure gradient field and an event node distribution diagram through obtaining the transportation vibration energy; if a creep coefficient resetting algorithm is triggered, mapping the event node graph into a weighted directed graph, extracting a key conduction path, and positioning an unpacking release type drift node; and introducing an unpacking stress correction term expansion cross-correlation entropy model, and determining a priority compensation strategy in combination with a time window information entropy to realize dynamic correction of sensor drift. The system comprises a hierarchy analysis module, a topology construction module, a lag positioning module and a drift compensation module. Through cross-scale modeling of vibration energy, material creep and drift compensation, the problem of sensor reading lag caused by stress accumulation in a supply chain can be reduced, and the packaging optimization decision efficiency is improved.

The invention relates to the technical field of artificial intelligence, can be applied to business scenes such as financial science and technology and medical health, and discloses a multi-agent communication and decision-making method, device, equipment and medium. The method comprises the following steps: obtaining state information and communication demand information of each agent in a multi-agent system, determining a communication weight based on the information, a dynamic communication topology is constructed in the system, an intelligent agent is used as a graph node, information of neighbor nodes in communication connection with the node is aggregated to generate interactive feature information, a corresponding decision strategy is generated by combining state information of the intelligent agent and the interactive feature information, and joint modeling and optimization of communication and decision are achieved. According to the method, dynamic topology construction is driven through communication weight, interactive feature information is introduced in strategy generation, and a collaborative optimization mechanism is formed by communication and decision, so that redundant communication is reduced, the communication efficiency is improved, and the multi-agent collaborative precision and the task execution effect are enhanced.

The invention relates to the technical field of wireless communication, and discloses a method and a device for constructing a wireless ad hoc network without a signal area based on chain networking, and the method comprises the steps: carrying out the collaborative fusion of an identity label and a communication capability of a node in the chain networking, and obtaining the node capability information of the chain networking; performing relay capability evaluation on a target node in the wireless ad hoc network to obtain a topology construction basis; performing optimization forming on candidate nodes in the wireless ad hoc network to obtain a chain network topology of the wireless ad hoc network; establishing a bidirectional communication link between nodes in the chain network topology to form a preliminary communication network of the wireless ad hoc network; detecting the link communication quality of the preliminary communication network step by step to obtain a hierarchical link state of the wireless ad hoc network; dynamically negotiating a hierarchical link state, and establishing a stable communication path of the wireless ad hoc network; according to the invention, the construction accuracy of the chain networking-based wireless ad hoc network in the non-signal area can be improved.

The invention discloses an OHT track node automatic identification and path planning method and system, and relates to the technical field of OHT tracks, and the method comprises the following steps: identifying track node positions based on vehicle motion track feature analysis, point cloud geometric clustering processing and a multi-mode verification method combining visual and magnetic features; performing semantic recognition according to the node positions to obtain node types, and constructing a track topological graph according to the node types; a candidate path set is generated through the track topological graph, and local optimization is conducted on the task execution sequence of the candidate path set and the vehicle matching relation; based on a local optimization result, performing synchronous adjustment on path selection and task allocation, and outputting an integrated scheduling scheme; according to the method, high-precision node positioning, semantic classification, structured track topology construction and integrated optimization of multi-vehicle path and task scheduling can be realized, and the operation efficiency, the scheduling reliability and the resource utilization rate of an OHT system are remarkably improved.

The invention provides a hierarchical clock tree synthesis method and system based on a hybrid topological structure, and aims to collaboratively optimize clock skew and clock delay in a chip clock network. The method comprises the following steps: firstly, introducing a balanced greedy search clustering algorithm, clustering registers at the initial stage of clock network construction, and reducing local clock skew and delay; and generating a bottom layer clock tree in each cluster by adopting a bounded deviation construction method so as to ensure the clock synchronization quality in the cluster. In a top-layer clock tree construction stage, the invention provides a self-adaptive H-tree topology construction algorithm, and the H-tree structure is dynamically adjusted according to the position relationship among all layers of clusters, so that a global clock distribution path is optimized under multiple layers. Besides, in order to further improve the CTS quality, the invention further designs a clock network optimization strategy based on buffer post-processing, and local buffer insertion and position adjustment are carried out for the problems of fan-out unevenness and load overrun so as to eliminate potential illegal paths.

The invention relates to the technical field of data processing, in particular to an enterprise right and interest management system based on a cloud platform, which comprises a right and interest topology construction module, a candidate node mapping module, a local conflict graph generation module, an enterprise right and interest management module, an enterprise right and interest management module and an enterprise right and interest management module, the weight iteration screening module locks the maximum weight value nodes to generate a locked node set, and the compatible combination generation module generates a right and interest combination list according to the set. According to the method, the mutual exclusion topology network is established by analyzing the original record, the flattened data is converted into the graph structure, the conflict graph is constructed to present the conflict state, the graph is traversed, the node with the maximum weight value is preferentially locked, and the connected conflict nodes are removed, so that the full-table scanning loss of the traditional database is effectively avoided; right and interest combination value maximization calculation under the complex mutual exclusion logic is achieved, and the right and interest distribution accuracy and the system processing efficiency are greatly improved.

The invention discloses a drainage system state space-time prediction method and system based on a neural network, and belongs to the technical field of computer system modeling and artificial intelligence, and the method comprises the steps: obtaining pipe network geographic structure data and operation observation data, carrying out the time marking, carrying out the event-driven space-time alignment and missing completion, generating an event alignment sequence, and carrying out the event-driven time-time alignment and missing completion; fusing the pipeline connection relation and the flow direction rule to construct a reachable topological graph; performing combined feature extraction on the event alignment sequence to generate a topological space-time representation tensor, applying constraints based on mass conservation and capacity limitation, and calculating and generating a physical constraint deviation value; and fusing the topological space-time representation tensor and the physical constraint deviation value to generate a space-time prediction field, and carrying out secondary mapping based on uncertainty quantization and error feedback to generate an adaptive correction field. By integrating event-driven data alignment, physical topology construction and constraint fusion learning and combining error correction, space-time prediction of the future state of the drainage system can be realized.

The invention relates to the technical field of display, and discloses an LED screen correction coefficient adjusting method and system, and the method comprises the steps: collecting the original brightness and chromaticity data of each LED unit; constructing a global topological connection relation graph; calculating an initial correction coefficient; the corrected residual error of the boundary pixels of the adjacent units is analyzed; establishing a consistency constraint equation set aiming at minimizing the difference between the boundary brightness and the chromaticity; and performing joint solution to obtain a global optimization correction matrix and writing the matrix into a driving module. The system comprises a data acquisition module, a topology construction module, an initial correction module, a residual analysis module, a constraint construction module, a global optimization module and a coefficient writing module, and integrates a temperature drift compensation and visual verification mechanism. Through cross-unit collaborative optimization, seamless fusion of splicing areas is realized, and brightness uniformity, chromaticity consistency and long-term operation stability are improved.

The invention discloses a micro-service fault diagnosis method and system based on a large language model and a knowledge graph. The method comprises the steps that a service dependency graph is automatically generated through static topology construction; converting the heterogeneous log into a standardized entity relationship triple by utilizing a semantic fusion engine; a diagnosis sub-graph associated with attributes is created through a dynamic sub-graph generation mechanism; and a workflow inference engine is adopted to realize multi-attribute collaborative root cause positioning, and a diagnosis report containing fault root causes and diagnosis results is generated. According to the method, operation and maintenance personnel can realize cross-service, high-fidelity and self-adaptive fault diagnosis in a fragmented observation data environment, and the diagnosis efficiency and accuracy are remarkably improved.

The invention discloses a server fault diagnosis system based on service topology correlation analysis, and relates to the technical field of computer data processing and automatic fault localization, and the system comprises a service topology construction module which is used for collecting multi-source heterogeneous data to construct infrastructure, runtime environment and business performance feature vectors, a service dependence topological graph is maintained; the fault storm diagnosis module is used for identifying a fault storm by calculating node state weights and utilizing spatial compactness and a time synchronization coefficient, and positioning a root cause node based on reverse breadth-first search and a fault propagation index; and the fusion alarm generation module is used for calculating the response time contribution degree of the root cause node to verify the causal relationship so as to distinguish dominant storm and hidden faults. The system executes cascade alarm convergence for dominant storm, triggers snapshot capture during deep operation for hidden faults, locates hidden root causes under complex topology, and realizes automatic fault closed-loop processing.

The invention relates to the technical field of computer hardware resource management, in particular to a dynamic load balancing optimization system for a computer mainboard, which comprises a data acquisition module, a feature decoupling module, a load evaluation module, a topology construction module and a strategy generation module. The system collects real-time load data of hardware components of a mainboard, and calculates intensive, data throughput and input and output blocking load features through feature decoupling separation. And generating a load pressure index and a migration potential coefficient of each component based on the decoupling characteristics, and dynamically constructing a mainboard resource topological graph. The system simulates the influence of different load migration schemes on delay and energy consumption on a topological graph, and generates a collaborative optimization instruction set integrating task scheduling, frequency adjustment and power distribution strategies. According to the method, fine identification of load characteristics and global collaborative scheduling of system resources are realized, and the resource utilization efficiency and the energy efficiency ratio are improved.

The invention relates to the field of engineering consultation service quality control, and discloses a service quality evaluation system, which comprises a data acquisition module, a dynamic topology construction module, a dual-channel evaluation module, an improvement suggestion generation module and an evaluation report generation module. A dynamic topology network with engineering documents as nodes and multi-dimensional association as edges is constructed, deep semantic matching of a real-time compliance detection channel and quality state differential equation prediction of a deep evolution analysis channel are combined, and standard compliance quantitative evaluation and quality evolution trend analysis are achieved. The abnormal propagation path detection based on the topology gradient field is linked with the industry knowledge graph to generate optimization suggestions, and the multi-source evaluation data is integrated into a standardized report through a dynamic association mapping technology. According to the method, the structured, dynamic and intelligent levels of engineering consultation service quality evaluation are improved, and the method has good practicability and engineering adaptation capability.

The invention discloses a power grid dynamic topology generation system based on machine account and real-time operation data fusion, and belongs to the technical field of electric power information. Comprising a historical data acquisition module used for acquiring historical topological data and real-time operation data; the rule base generation module is used for establishing a graph construction rule base; the data acquisition module is used for acquiring machine account data and real-time operation data of a target power grid; the power grid layering and grading module is used for carrying out layering and grading processing on the target power grid; the partition topology construction module is used for generating local power grid topology; and the topology merging module is used for merging the local power grid topologies. According to the method, by establishing the rule base and combining the technical means of hierarchical processing, feature matching and the like, the problems that in the prior art, the power grid topology construction speed is low, and the adaptability to a large-scale power grid is insufficient are solved, and efficient and accurate power grid dynamic topology generation is achieved.

The invention relates to the field of space intelligence, and provides a cross-scene space intelligence semantic topology generation method and system, and the method comprises the steps: introducing a transition region modeling mechanism with a transition node as a core in a cross-scene space semantic topology construction process, regions which are mutated at the same time in geometric structures, semantic attributes and topological scales among different space scenes serve as independent objects to be recognized, expanded and processed, and the problems of semantic fracture, topological distortion and space incoherence caused by simple splicing only in the scene boundary level in the prior art are solved. Transition nodes are subjected to structure expansion based on the topological connection relation, a transition area reflecting the communication relation of different space scenes is constructed, and semantic transition consistency constraint and topological structure deformation compensation are implemented in the transition area. In addition, in combination with a multi-scale topology adaptive adjustment mechanism, the cross-scene semantic topology can dynamically adjust the node granularity and the topology level under different spatial scales.

The invention discloses a digital twin-driven intelligent construction site resource dynamic scheduling system and method, and belongs to the technical field of building construction management, and the system comprises a four-dimensional resource topology construction module which constructs a four-dimensional topological graph of construction resources based on UWB and RFID technologies; the mixed engine scheduling module adopts a mixed integer programming engine to generate a reference scheduling scheme in an initial stage, and adopts a reinforcement learning optimization engine to perform dynamic adjustment in an execution stage; the conflict prediction network module is used for establishing a conflict prediction model based on historical data and triggering a resource redistribution mechanism; the virtual interaction visualization module displays a scheduling scheme in a virtual reality environment and supports manual intervention, depth relation modeling, double-engine collaborative optimization and active conflict prevention of construction resources are achieved, the resource utilization rate is increased by 35% or above, the construction period is shortened by 15% or above, and conflict events are reduced by 50% or above.

The invention provides a large-scale low-orbit constellation topology construction method and system based on the reliable reachable distance of an optical transceiver and a storage medium, and the method comprises the steps: 1, building an inter-satellite laser link probability evaluation model based on the capturing probability of an aiming-capturing-tracking system and the motion characteristics of satellites; 2, on the basis of the inter-satellite laser link probability evaluation model, constructing a giant satellite network topology evaluation model with inter-satellite link time average stability, satellite hop count and communication time delay as core optimization targets, and outputting a globally optimized laser link connection scheme; and 3, for a laser link connection scheme, proposing a giant constellation network topology design algorithm based on a capture probability, and constructing a constellation topology by adopting a minimum unit of a space symmetric constellation. The method has the beneficial effects that the constellation topology is constructed by adopting the minimum unit of the space symmetric constellation, so that the topological structure is stabilized, the channel gain is improved, the end-to-end delay is shortened, and the average hop count and the variance of the network are reduced.

The invention relates to the field of business process optimization, in particular to a business process intelligent reconstruction optimization method and system. The intelligent reconstruction optimization system for the business process comprises a topological structure construction module, a causal value analysis module, an optimization instruction generation module and a reconstruction scheme generation module. According to the invention, through collaborative analysis of a causal discovery algorithm and a dynamic value quantification model, precise targeted intervention of business process optimization is realized for the first time; in causal analysis, interference of confounding factors is effectively stripped based on an anti-fact verification mechanism of interference grouping variables, and root cause active nodes causing performance bottleneck are accurately identified; in value analysis, value contribution scores are calculated by fusing user feedback emotion polarity and conversion rate data, three types of value attributes are dynamically marked, and resource mismatching caused by single-dimension optimization in a traditional scheme is overcome.

The invention discloses a fully mechanized coal mining hydraulic support electrohydraulic control system based on deep learning, and the system comprises a data collection and topology construction module which is used for collecting multi-source real-time data and constructing a support topology adjacency matrix; the data preprocessing and dynamic topology updating module is used for preprocessing the original data set and dynamically updating the initial topology adjacency matrix; the space-time diagram convolution depth autoregression modeling module is used for outputting a future multi-step prediction result set based on a space-time diagram convolution depth autoregression network model; the underground safety constraint judgment module is used for performing underground safety constraint judgment; the control instruction generating and issuing module is used for generating an electro-hydraulic control instruction set; and the execution feedback and model closed-loop updating module is used for collecting real-time state data, updating the preprocessed data set and enhancing the spatial-temporal characteristic sequence. According to the invention, the space-time diagram convolution depth autoregressive network is adopted to realize intelligent control of the fully-mechanized mining hydraulic support.

The invention relates to the technical field of multi-unmanned aerial vehicle geological surveying and mapping, and discloses a geological surveying and mapping method and system based on multi-unmanned aerial vehicle cooperation. A distributed topology construction unit of the system constructs a directed graph topological structure according to a multi-unmanned aerial vehicle cluster communication relationship, defines an in-neighborhood node set and an out-neighborhood node set and initializes a state observer; the geological data acquisition unit controls the unmanned aerial vehicle to fly according to a preset path and acquires geological deformation, environment temperature and humidity and soil compactness data to generate an original data set; the time-space analysis unit analyzes the timestamp to generate a dynamic time axis and marks geological event nodes; the dynamic environment correction unit calculates an environment correction factor and corrects geological deformation data to generate an environment adaptive data set; and the collaborative state observation unit inputs the adaptive data set into an observer, and generates a local geological state estimation result by combining with a neighborhood node estimation value update parameter. The system can improve the accuracy of geological surveying and mapping data and the precision of state estimation, and is adaptive to a dynamic working environment.

The invention provides a power distribution network topology construction splicing method based on recursive grouping and a graph attention network. The method comprises the following steps: acquiring electrical attribute information of node equipment in a power grid by using a monitoring sensor; recursive grouping is carried out, nodes and branches in the feeder line of the power distribution network are identified by using the constructed impedance matrix and distance matrix, an initial topological graph is constructed, and a node feature matrix and an initial adjacent matrix are obtained; a GATv2 graph attention network is introduced, dynamic attention calculation is performed by using the node feature matrix and the initial adjacent matrix, and a final topological graph of the feeder is generated; a landmark matching method and Pearson's correlation coefficient calculation are adopted, and boundary points of each independent topology are identified and matched; performing accurate alignment on the topological graph generated by each feeder line by using a graph embedding synchronization technology, and finally performing global topology processing; and carrying out topology verification on the constructed global topological graph. Construction and efficient splicing of line topologies in different areas of the power distribution network are achieved, and efficient operation and intelligent development of a modern power distribution network are supported.

The invention relates to a flood early warning and valve linkage control system based on the Internet of Things. A river topology construction module dynamically generates a topology network representing a river structure by fusing geographic information and real-time sensing data; the early warning weight calculation module generates a quantitative risk index in combination with the water level overrun degree and the historical flood probability; the water flow time matrix generation module calculates a flood propagation time sequence based on a hydraulic evolution model; the flood early warning control module synchronously outputs an early warning report and a valve control instruction through a space-time risk coupling mechanism to realize intelligent linkage of early warning and control; the topology model correction module dynamically optimizes the river channel parameters and the risk model according to the valve execution feedback. According to the system, the problems of early warning response lag, model prediction misalignment and extensive valve control in traditional flood control are solved, early warning and control collaborative decision is driven through data closed loop of the Internet of Things, and the accuracy and reliability of flood control response are improved.

The invention belongs to the technical field of low-voltage transformer area topology identification, and particularly relates to an electricity-obtaining data-based electricity meter topology automatic hitching method and system. Comprising the following steps: acquiring an opening event of an intelligent tail-end circuit breaker through an intelligent fusion terminal of a low-voltage transformer area; the intelligent fusion terminal issues a closing instruction to the plurality of intelligent tail end circuit breakers in sequence according to a set time interval T; the intelligent fusion terminal receives the electricity obtaining time of the household meters reported after the household meters are electrified, and collects the closing time recorded by the intelligent tail end circuit breakers or the measuring switches; and carrying out clustering comparison on the electricity obtaining time of the household meters and the closing time recorded by the intelligent end circuit breakers or the measuring switches, and determining the hooking object of each household meter. According to the method, the electricity obtaining time of the household meters and the closing time recorded by the intelligent tail end circuit breakers or the measuring switches are subjected to clustering comparison, the hooking object of each household meter is determined, automatic hooking of the household meters is achieved based on the determined hooking objects, and the intelligence and automation degree of precise topology construction is improved.

The invention discloses a loose coupling unified task arrangement method supporting multi-dimensional expansion, which comprises the following steps of: firstly, defining a transmission agreement of task arrangement: constructing a unified intermediate representation structure for describing an arrangement process, the intermediate representation structure being used for recording a process topological structure, a node type and parameter configuration; secondly, on the basis of the intermediate representation structure, an executable directed acyclic graph is automatically constructed by adopting a child node matching construction algorithm based on queues and backtracking; and thirdly, executing an arrangement process based on the constructed directed acyclic graph: starting from all write nodes, recursively calling a data processing method of a father node until reaching a read node, and executing a complete data processing link in a bottom-up manner. Through a unified intermediate representation structure and a queue backtracking matching algorithm, automatic topology construction and dependency management of a complex process are realized, flexible expansion of node types is supported, development complexity is reduced, and adaptability, efficiency and maintainability of a data management system are improved.

The application provides a distributed remote sensing interpretation vector post-processing method based on a grid DCEL topology. The method comprises: performing grid slicing on global grid data to obtain grid data of a plurality of grid units; constructing a bidirectional edge connection table topology based on the grid data of each grid unit respectively; performing elimination processing, thinning processing and smoothing processing on the topology of each grid unit respectively; merging the topology of each grid unit to obtain a global topology; and exporting the global topology as vector data. In this way, a high-performance grid topology engine based on a distributed architecture can be provided, a DCEL topology structure can be constructed from grid data, and distributed computing can be supported through a topology level merging algorithm; the grid is used to construct the topology, thereby reducing the steps of grid vectorization, and the grid-based topology construction method can naturally support distributed computing.

The invention relates to a black box graph injection attack method and device based on a thermonuclear, and aims to solve the problems that the existing graph injection attack (GIA) is insufficient in imperceptibility, limited in application scene and high in calculation cost, and improve the effectiveness and concealment of a graph neural network (GNNs) attack. The method comprises the following steps: firstly, screening candidate nodes based on classification margins and node degrees, and grouping according to most likely misclassification labels; secondly, sampling high-frequency feature dimensions of similar original nodes to generate injection node features consistent with original image feature distribution; then, adopting a two-stage topology construction strategy: constructing a temporary perturbation graph in the first stage, and screening out an injection edge conforming to the topological characteristics of the original graph through node-level thermonuclear signature (HKS) camouflage constraint and graph-level thermonuclear matrix guidance in the second stage; and finally, generating an imperceptible perturbation graph, and reducing the classification performance of the target GNN model on the premise of not exposing attack behaviors. According to the method, the structure or parameters of the target model do not need to be obtained, the method only depends on the black box query and proxy model, the practicability, imperceptibility and calculation efficiency of attacks are remarkably improved, and the method is suitable for GNNs attack testing in safety key scenes such as the Internet of Things.