123 results about "Dependency network" patented technology

The invention discloses an emotion analysis method and system based on multi-modal fusion. The method comprises the following steps: S1, acquiring multi-modal data and initial feature representation of each modal; s2, inputting the initial feature representation of each mode into a multi-head attention mechanism for preliminary fusion, and generating a primary fusion feature; s3, in combination with the bidirectional long-short-term memory network and the multi-scale causal convolutional network, generating time sequence features; s4, constructing a hierarchical dependency relationship network among modals by using a multilayer graph neural network, and forming global context sensing feature representation; s5, dynamically generating and adjusting the fusion weight of each mode; s6, combining classification and regression output by using a multi-layer full-connection network and a multi-task learning framework, and generating multi-dimensional expression of an emotional state; and S7, migrating the group sentiment model to specific sentiment analysis of individual users through transfer learning and personalized modeling technologies. According to the method, high-precision and multi-dimensional sentiment analysis is realized by utilizing multi-modal fusion and self-adaptive weight adjustment.

The invention discloses a communication equipment fault intelligent diagnosis method and device, equipment and a medium. The method comprises the steps of collecting equipment operation logs in real time, and extracting an abnormal point set; clustering the abnormal points, and matching a hardware fault, a software error and a network congestion type in combination with a preset mode library; dimensionality reduction is carried out on temperature, voltage and current sensor data based on principal component analysis, and a time-synchronized fusion feature vector is constructed; analyzing features and abnormal modes, and outputting potential fault points; associating the classification model to generate a fault priority by extracting the dynamic characteristics of the network flow; and a dependency network is constructed in combination with Bayesian reasoning, so that accurate fault positioning is realized. The system supports incremental learning, updates an abnormal mode library and automatically expands a knowledge base. Through multi-dimensional data fusion and spatio-temporal feature joint modeling, the problems that a traditional method depends on manual rules and is high in false alarm rate are solved, and the method is suitable for complex scenes such as a 5G base station and a data center.

The invention relates to the field of servo motor parameter optimization, in particular to a magnetic field simulation and parameter optimization method and system for a permanent magnet servo motor rotor. The method comprises the following steps: acquiring a rotor global magnetic field monitoring curve in the operation process of a servo motor rotor, and constructing a global magnetic field accurate compensation model; performing regional magnetic circuit structure analysis on the global magnetic field accurate compensation model, performing dynamic magnetic field logic mapping modeling, and constructing a real-time regional magnetic field logic map; performing magnetic field state change analysis on the global magnetic field accurate compensation model, performing multi-node change track correlation analysis, and constructing a multi-node magnetic field coupling dependency network; and performing short-term magnetic field evolution prediction based on the multi-node magnetic field coupling dependency network, and performing comprehensive electromagnetic performance evaluation to obtain an electromagnetic performance evaluation report. According to the method, the sudden change of the magnetic field is accurately and quickly identified, so that the magnetic field parameters are iteratively adjusted and optimized, and the service life and the performance of the servo motor are improved.

The invention provides a knowledge graph construction method and device for table data relations, equipment and a medium, and belongs to the technical field of knowledge graphs. The method comprises the steps that when a user triggers cell operation, a system automatically captures a cell reference formula, a working page where the cell reference formula is located and coordinates; and through an analysis formula, identifying the dependency relationship of the cells, and positioning the associated cells. Checking whether the associated cell contains a reference formula, if so, regarding the associated cell as a new current cell, and repeating the analysis process; and if not, creating a tracking relation knowledge graph by utilizing the working pages, the coordinates and the dependency relation of the current cell and the associated cell, and displaying the dependency network among the cells in a graphical mode. According to the method, automatic analysis of the cell formula is achieved through an intelligent means, a clear data association graph is constructed, the tracking efficiency of the complex data dependency relationship is greatly improved, and the full view relationship of the data is visually displayed.

The invention discloses a nuclear accident emergency three-layer coupling network model and a construction method, and relates to the crossing field of nuclear accident emergency management and complex system risk simulation. The nuclear accident emergency three-layer coupling network model comprises an RMN layer which is used for simulating the dynamic diffusion process of radionuclides leaked by a nuclear accident in the atmosphere; the IN layer is used for constructing a spatial distribution and dependency relationship network of infrastructures; the EN layer is used for topologically defining a command or cooperation relationship of the emergency organization; according to the three-layer network coupling mechanism, when RDR of a middle grid is larger than or equal to RDRmax-1, IN-layer nodes in the middle grid at the same position are marked to be in a damaged state; and quantifying the bidirectional dependence density of the IN layer and the EN layer through the coupling strength epsilon. The method is used for predicting the direct risk of radioactive contamination in a nuclear accident and the secondary risk caused by emergency evacuation, revealing the risk evolution law, quantifying the risk diffusion scale, determining key influence factors, and providing a theoretical basis for formulating an emergency strategy.

The invention relates to the technical field of network security detection, and discloses a network transmission protocol intrusion detection method and system. The method comprises the following steps: acquiring a network transmission protocol data stream, extracting head features and load features of a data packet, and integrating to form an original feature set; the original feature set is classified according to protocol types, different protocol clusters are divided, and corresponding identifiers are marked; constructing a dynamic feature filter based on the protocol cluster identifier, and screening a feature subset associated with the current protocol cluster from the original feature set; inputting the feature subset into a protocol state analysis model, and outputting a protocol state vector and abnormal behavior probability distribution by the model; and generating an intrusion detection result according to the two, and triggering a corresponding defense response instruction. According to the method, different protocol characteristics are adapted through protocol cluster classification and dynamic characteristic screening, accurate identification of protocol intrusion behaviors is realized, and the method can be applied to various scenes depending on a network transmission protocol.

The invention provides a cloud native micro-service splitting and reconstructing method based on runtime behaviors. A three-layer architecture design is adopted, and the system comprises a dependency network graph construction module, a micro-service division module and a code reconstruction module. The dependency network diagram construction module models specific project source codes and user access operation into abstract representation capable of being processed by an algorithm, the micro-service division module divides a network diagram, obtains a splitting scheme for the dependency network diagram through visual page interaction with a user and manual modification, and inputs the splitting scheme into the code reconstruction module; and a code reconstruction module performs code reconstruction based on the splitting scheme and the dependency network diagram to obtain a complete and operable micro-service architecture code, so that the effect of automatically analyzing Spring Boot project source codes and monitoring the running project to generate the dependency network diagram is realized.

The invention discloses a medicine sales management method and system based on information monitoring, and relates to the technical field of medicine sales management, and the method comprises the steps: collecting medicine sales time sequence data, carrying out the cleaning and structural processing, and generating standardized sales data; based on the standardized sales data, constructing a sales behavior time sequence causal network, embedding virtual sales intervention nodes, and generating a time sequence causal graph; the intervention strategy is coded to the virtual intervention node in the time sequence causal atlas, the graph neural network is utilized to simulate time sequence propagation of the intervention strategy, and a sales prediction result is obtained; and combining the sales prediction result with the real-time inventory, constructing a multi-level distributed sales dependency network, and forming the sales dependency network. According to the method, the preset intervention strategy is embedded into the causal structure in the node form, so that dynamic modeling of the multivariable causal relationship in the sales behavior is realized, and the prediction result not only reflects the historical trend, but also can respond to the possible influence of different intervention strategies.

The invention discloses a data transparent encryption and decryption method and system based on cloud authentication, the system comprises an encrypted file server, a cloud server and an electronic device, the cloud server deploys a cloud authentication system, the cloud authentication system comprises a cloud authentication platform and a UTOFS file system running on the electronic device; the cloud authentication system performs user creation, authentication and key generation, the electronic device logs in the cloud authentication system, authenticates and mounts an encrypted file, and performs file encryption and decryption based on the key; the encryption and decryption process is realized on a system layer of the computer, binary files are used, the data protection level is high, leakage of the files is avoided from the source, and data cannot be read normally after the computer leaves an authorized environment; meanwhile, the encryption and decryption process does not depend on the network to carry out transparent encryption and decryption, and the real-time performance of data encryption and decryption is guaranteed.

The embodiment of the invention discloses a component recommendation method and device, electronic equipment and a storage medium, and the method comprises the steps: responding to a component recommendation operation triggered by a target user; obtaining a component dependence network graph, component demand description information of the target user and historical interaction information of each historical interaction component used by the target user in the current editing item; determining a plurality of to-be-screened components from a component library on the basis of the component dependency network graph and the historical interaction information; based on the component dependency network diagram, the component attribute information of each component in the component library, the component demand description information and each piece of historical interaction information, determining the adaptation attribute of each component relative to the target user; according to the method, the to-be-screened components are screened on the basis of the adaptive attributes, the target recommended component is obtained, the component search efficiency and modeling fluency are remarkably improved, the burden of repeatedly trying different component search conditions by a user is reduced, and finally the efficiency and quality of the overall modeling process are optimized.

The invention discloses an AI test case intelligent optimization verification method and system, and the method specifically comprises the steps: employing a causal discovery algorithm to analyze an implicit dependency relationship for an entity association structure in a test case knowledge graph, recognizing redundant test cases and invalid test paths, and obtaining a causal dependency network; on the basis of a causal dependence network, traversing uncovered entity relationship paths in the test case knowledge graph, and generating supplementary test case suggestions; dynamically adjusting analysis strategy parameters according to historical optimization data through a self-adaptive engine to obtain an optimization test case set; and based on the optimized test case set, generating an input parameter variation sample by using a fuzzy test engine, and expanding a test boundary. According to the method, the black box test efficiency can be remarkably improved, the test coverage rate and accuracy are improved, meanwhile, the workload of a test engineer is relieved, the user experience is improved, and an efficient and intelligent solution is provided for the field of software test.

The invention discloses a method based on Bayesian ATTamp; the invention discloses a general complex network attack inference method for a CK network, and relates to the technical field of network security. Performing sequential logic modeling on tactics in the CK framework to ensure that a generated attack sequence is logically coherent and reflects a behavior pattern of a real complex network attack; based on the tactical sequential logic sequence, guiding of expert knowledge on tactical transfer is expanded into guiding on technology transfer, and a technology dependence network is constructed and formed; an anonymized data set from a real network attack event is adopted, and a technical dependency network is further optimized by utilizing a technical co-occurrence relation in the data set; bayesian ATTamp is carried out on the basis of a technical co-occurrence relation between a technical dependency network and a real network attack data set; the CK network performs structure learning and parameter learning, constructs a network model independent of a specific threat detection system, and realizes complex network attack inference independent of a specific system.

The present application discloses a method for finding useless files. First, find the paths and names of all files according to the directory index file. Then, divide all files of the software development project into two categories according to the file suffixes: code files and resource files. Then determine and record the dependencies of each code file. Then merge the dependencies of all code files to generate a complete dependency network of all code files. Finally, in the complete dependency network of all code files, the files corresponding to the independent subnets that do not contain the main file are regarded as useless files. The present application first forms a dependency network between files, and filters out useless files in combination with filtering rules, so that the search results are accurate and comprehensive.

The present invention discloses an autonomous positioning and mapping method for an inspection robot based on semantic information of a plant or station, comprising data annotation of plant or station images, construction of a specific semantic image dataset applicable to a variety of image segmentation and target detection algorithms, combination of target detection, optical flow motion consistency detection and ORB feature extraction, and consideration of pure vision and visual odometer schemes combining vision with IMU, thereby maximizing the accuracy of the inspection robot in calculating its own posture in dynamic scenes, integrating a dual-channel coordinate attention mechanism into a real-time semantic segmentation network BiSeNet to optimize its network structure, and mapping the image segmentation results into the ultimately generated point cloud, thereby completing the construction of a semantic point cloud map and a semantic octree map and realizing map reuse. The present invention reduces the inspection robot's dependence on the network, so that its positioning and subsequent path planning tasks no longer need to rely on network maps.

According to the active power distribution network adaptive stage difference cooperation method based on positive sequence voltage spatial distribution, the problem of traditional protection setting failure caused by DG random access and network topology flexible adjustment is solved, and the spatial characteristics that positive sequence voltage fault components are distributed along a feeder line and are strongly related to fault positions are utilized; constructing a logarithmic inverse time limit characteristic equation based on the component; and by comparing actual measurement values and reckoning values of local and adjacent protection positive sequence voltage fault components in real time, online self-adaptive setting of equation parameters is carried out. And on the basis of parameters updated in real time, adjacent protectors calculate respective action time in parallel, and an action time level difference meeting absolute selectivity is dynamically generated, so that the remote backup protection quickness is effectively improved. According to the method, the starting criterion and the steady-state voltage action criterion for distinguishing the fault and the disturbance are combined, and finally self-adaptive step-by-step matching of multi-stage protection independent of network topology and power supply types is achieved.

Disclosed is a method of constructing long and short-range dependency network learning model: that includes obtaining heart sound data having plurality of audio files; preprocessing the plurality of audio files; extracting Mel-Frequency Cepstral Coefficients (MFCCs) from the preprocessed plurality of audio files; restructuring extracted MFCCs into an input layer of the long and short-range dependency network learning model; implementing two or more LSTM network layers with a dropout; and employing a softmax activation function to a final output layer. Disclosed also is a method of classifying heart rate sound data.

The invention discloses a software multi-defect positioning method and device based on variable dependency and clustering. According to the method, for a program block covered by a failed test case, a static analysis method is adopted, a data dependency, control dependency or function dependency relationship among variables is extracted, and a variable dependency network model is constructed to represent defects. Based on the above defect characterization, a graph editing distance is used to measure the similarity between the different characterization. And then selecting a proper clustering algorithm to adapt to fault characterization and similarity measurement modes, and automatically grouping failed test cases according to defect types to be used for positioning. According to the method, a more comprehensive and effective defect characterization algorithm is constructed, and a more effective failure test case difference measurement method is combined, so that the number of correctly predicted defects is effectively increased. By improving the accuracy of defect clustering, an independent defect focusing suspicious degree ranking is generated, and an effective solution is provided for a multi-defect positioning problem.

The invention provides an equipment asset management system and method based on a graph model, a data module is used for acquiring sample data, and the sample data comprises network equipment in a network service architecture and a dependency relationship among the network equipment; the graph model module is used for constructing a graph model according to the sample data, each network device in the network service architecture is a vertex of the graph model, and a dependency relationship among the network devices is an edge of the graph model; and the query module is used for receiving a query request of a user, performing query in the graph model by utilizing a graph query language based on the query request, and displaying a query result, and the query result displays the equipment dependence network with potential influence. By constructing the graph model, the dependency relationship between the devices can be effectively judged, and potential influence analysis is carried out, so that the device management efficiency is improved, and the device management effect is optimized.

The embodiment of the invention provides a component repairing method, system and equipment based on a software supply chain, a medium and a product, and relates to the technical field of information security. The method comprises the steps of obtaining component information of a software supply chain based on a full life cycle of software development; wherein the component information at least comprises a component material list, component dependency relationship data and component risk data; constructing a dependency network graph based on the component information; determining a graph type and a spectral radius of the dependency network graph; determining a propagation threshold corresponding to the graph type; if the spectral radius is greater than or equal to the propagation threshold, determining a to-be-repaired component set from the dependency network graph based on the graph type, the spectral radius and the propagation threshold; and repairing the to-be-repaired components contained in the to-be-repaired component set to obtain a repaired software supply chain. The security defense capability of the software supply chain can be improved, and stable and reliable operation of a software system is guaranteed.

The invention relates to an unmanned aerial vehicle cluster communication network full-end reliability prediction method based on GNN and network signature, and belongs to the technical field of communication networks, and adopts two-stage prediction: a first stage, determining topological structure characteristic variables of an unmanned aerial vehicle cluster communication network, generating a training data set based on simulation, and designing, training and optimizing a graph neural network model; the prediction of network signature vectors is realized, and the signature only depends on network structure characteristics and does not depend on link reliability parameters; and in the second stage, on the basis of the signature vector, the all-end reliability of the network in a link failure scene is calculated by combining link reliability parameters and utilizing a signature-based all-end reliability deterministic hybrid expression. According to the method, the sensitivity of the model to the reliability parameter change is reduced, and the generalization ability and applicability of the prediction method under different network scales and reliability conditions are improved.

The application provides a multi-event extraction method, system, device and medium based on a graph self-dependent network, relates to the field of natural language processing information extraction, and the method comprises the following steps: acquiring primary prediction information; adopting a graph self-dependent network to perform multi-event dependency relationship modeling to obtain global event information; outputting trigger word text based on the primary prediction information and the global event information; adopting a conditional fusion function to integrate the trigger word text in the detection text to obtain the primary prediction information; adopting the graph self-dependent network to perform multi-argument dependency relationship modeling to obtain global argument information; and outputting event argument text based on the primary prediction information and the global argument information. The application uses the graph self-dependent network to mine the dependency between multi-events, does not need to rely on other natural language processing tools, allows learning the relevance of different trigger words and arguments, and makes the graph self-dependent network have good expansibility and effectiveness.

The invention discloses an automatic deployment method for a long safety chain network based on Kubernetes Operator, and the core scheme of the method is to define double-layer custom resources (CRD): ChainMakerNetwork is used for declaring chain-level global configuration, ChainMakerNode is used for declaring node-level configuration, through the cooperative work of a node controller and a network controller, the network controller is used for declaring the global configuration of the chain-level, and the network controller is used for declaring the global configuration of the chain-level and the node-level configuration of the chain-level. A node controller generates an identity for each node and reports the identity to a resource state field in advance, and a network controller aggregates information to generate a global chain configuration file after monitoring that all dependencies are ready, so that the mechanism of the scheme breaks the circular dependency that node starting depends on network configuration, and network configuration generation depends on node identity; according to the invention, one-key and declarative deployment and full-life-cycle management of the long safety chain network in the Kubernetes environment are realized, and the deployment efficiency, reliability and operation and maintenance automation level are greatly improved.

The present invention discloses a general complex network attack inference method based on the Bayesian ATT&CK network, which relates to the field of network security technology. Using the kill chain and empirical knowledge as the research basis, it conducts a temporal logic modeling of the tactics in the ATT&CK framework to ensure that the generated attack sequences are logically coherent and reflect the behavior patterns of real complex network attacks; based on the tactical temporal logic sequence, it extends the guidance of expert knowledge on tactical transfer to the guidance of technique transfer, and constructs a technique dependency network; it uses an anonymized dataset from real network attack events and further optimizes the technique dependency network by utilizing the co-occurrence relationship of techniques in the dataset; based on the technique dependency network and the co-occurrence relationship of techniques in the real network attack dataset, the Bayesian ATT&CK network performs structure learning and parameter learning to construct a network model independent of specific threat detection systems, realizing the inference of complex network attacks independent of specific systems.

This invention discloses a method, apparatus, electronic device, and storage medium for cross-device NPU invocation. The method includes: converting the NPU of a terminal device within a network into an NPU service object through an NPU virtual service, wherein the number of terminal devices within the network is at least two, and each terminal device within the network possesses an NPU; registering the NPU service objects of each terminal device to the distributed hardware management service by calling the registration interface of the distributed hardware management service through the NPU virtual service; when a target terminal within the network receives a task to be processed, querying the target information of each NPU service object within the network through the distributed hardware management service, and determining the target NPU service object for processing the task based on the target information. The technical solution of this invention solves the problems of current edge-cloud combined methods for processing neural network tasks, which rely on network latency and cannot meet all operating conditions, achieving efficient utilization of the NPUs of devices within the network and improving task processing efficiency.

This invention discloses a BIM model-based intelligent management method and system for prefabricated buildings, belonging to the field of prefabricated building management technology. Specifically, it includes: acquiring BIM data containing component geometric attributes, material properties, and installation logic relationships, as well as supply chain data on component production status, transportation location, and warehousing information; constructing an assembly dependency network with components as nodes and installation dependencies as edges; monitoring supply chain data in real time and identifying abnormal components; locating abnormal nodes in the assembly dependency network and traversing to determine the affected assembly paths; retrieving matching alternative components based on warehousing information and replacing nodes; and finally recalculating the installation sequence of affected components based on the updated assembly dependency network to generate a new assembly sequence. This method enables rapid response and dynamic adjustment of the assembly process to supply chain anomalies.

The present invention discloses a method for diagnosing network faults of an electric data center based on a dependent network model, including: constructing a cascading failure model of the dependent network of the electric data center according to the log information of the terminal device accessing the electric data center; calculating the initial load and capacity for each edge in the network, simulating random network failures and updating the network status, calculating the current load of each edge in the updated network, and deleting the edge from the network if the current load is greater than the capacity of the edge; repeating status updates and load calculations and deletions until all edges in the electric data center network meet the condition that the current load is less than or equal to the capacity of the edge, or all nodes in the network are isolated nodes; calculating the current network model function value, and providing a quantitative judgment result. The present invention is oriented to complex electric data center scenarios, and captures the real-time changes of node status through dependent network modeling, effectively supporting status perception and fault analysis.

The invention discloses a crack detection method and system based on edge end equipment, and the method comprises the following steps: collecting an image frame of a crack condition to be detected, and carrying out the preprocessing of the image frame; performing matrix construction on the preprocessed data, and performing image feature extraction on the preprocessed data by using convolution calculation; performing quantization processing on a feature extraction result by using an INT8 quantization algorithm; inputting the quantized data into a Leaky ReLU function to carry out data nonlinear transformation; constructing a 2X2 data matrix, and processing the data after nonlinear transformation by using a maximum pooling method; carrying out fusion operation on the feature maps of different levels; performing further feature extraction on the fused feature data to obtain output layer data; analyzing the output layer data; and converting the target frame information into RGB data for visual display. The target detection effect can be improved, and the problems of large model size, large equipment weight, high system power consumption, dependence on a network and the like in the existing field environment crack detection technology are solved.

This application discloses a diagnostic control method, apparatus, electronic device, and storage medium. When a first connector receives a first data frame that a first device needs to send to a second device, and the response type of the first data frame is a target response type, the first connector replaces the second device in sending a first acknowledgment frame for the first data frame to the first device. Specifically, after sending the data frame of the target response type, the first device interrupts the transmission of the data frame to be sent to the second device. Upon receiving the first acknowledgment frame, the first device continues to send the data frame to be sent to the second device to the first connector, so that the first connector can receive and forward the data frame to the second device. Based on this, by using the first connector, which is adjacent to the first device, to replace the second device in acknowledgment responses, the data frame transmission process does not require reliance on network acknowledgment interactions, reducing timeout problems caused by network latency.

The application relates to the technical field of network security detection, and discloses a network transmission protocol intrusion detection method and system. The method comprises the following steps: collecting network transmission protocol data streams, extracting packet header features and payload features, and integrating to form an original feature set; classifying the original feature set according to protocol types, dividing different protocol clusters and marking corresponding identifiers; constructing a dynamic feature filter based on the protocol cluster identifiers, screening a feature subset associated with the current protocol cluster from the original feature set; inputting the feature subset into a protocol state analysis model, outputting a protocol state vector and an abnormal behavior probability distribution from the model; generating an intrusion detection result according to the two, and triggering a corresponding defense response instruction. Through protocol cluster classification and dynamic feature screening, the method adapts to different protocol characteristics, realizes accurate identification of protocol intrusion behaviors, and can be applied to various scenes relying on network transmission protocols.

The invention discloses a response equipment adaptive dynamic regulation and control method under panel furniture production line disturbance. The method comprises the following steps: establishing a production line equipment dependence network model, defining equipment as nodes, defining a dependency relationship between the equipment as directed edges, and configuring weights; when a disturbance event is received, calculating an initial disturbance intensity value and determining an initial disturbance node; on the basis of the initial disturbance intensity value and the initial disturbance node, longitudinal propagation and transverse diffusion of disturbance in the network model are simulated, and a comprehensive disturbance index is calculated for each equipment node; comparing the comprehensive disturbance index with a dynamic response threshold value, determining a response equipment set, and outputting a regulation and control instruction according to the response equipment set; according to the method, through quantitative disturbance intensity and network propagation simulation, prediction and targeted intervention of the disturbance chain reaction are realized, the stability and efficiency of a production line in a dynamic environment are improved, and excessive response or insufficient regulation is avoided.