81 results about "Information dispersal" patented technology

The invention provides a carbon emission calculation method and device based on power grid prediction, and the method comprises the steps: constructing a graph structure containing a plurality of nodes and edges based on the topological information and operation data of a power grid; the nodes and the edges respectively represent the connection relationship between the devices in the power grid; on the basis of the graph neural network model, feature extraction and information spreading are carried out on the graph structure, a prediction result is obtained, and the prediction result comprises state information of each node; a physical constraint mechanism is arranged in the graph neural network model, and the physical constraint mechanism is used for constraining a prediction result to conform to a physical rule of power grid operation; and outputting a prediction result conforming to the physical rule of power grid operation, thereby solving the problem that the prediction result in the traditional scheme is unavailable and uncredible in the key industrial field, enabling the prediction result to be seamlessly connected to the power grid dispatching and decision-making process with extremely high requirements on safety and stability, and improving the reliability of the power grid. Therefore, the practicability of the intelligent prediction technology of the power grid is greatly improved.

The invention discloses an iterative collaborative learning method for node characterization and a topological structure, which comprises the following steps: acquiring initial graph data containing a node set and an initial edge set, and initializing node characterization; constructing a dynamic edge predictor, calculating the structure confidence degree between the node pairs based on the current node representation, generating a candidate edge weight matrix, and obtaining an optimized adjacent structure through a sparsification strategy; inputting the optimized adjacent structure into a graph neural network encoder, performing information propagation and aggregation in combination with the node feature matrix, and outputting enhanced node representation; the enhanced node representation is fed back to the dynamic edge predictor, and the edge weight and the node representation are iteratively updated until the convergence condition is met; and outputting a final optimized graph structure and node representation for downstream graph analysis. According to the method, through iterative collaborative optimization of the graph structure and the node representation, the graph topology can be dynamically corrected, the discrimination capability of the node representation is enhanced, the method is suitable for various complex scenes such as a noise graph and a sparse graph, and the performance of a downstream graph analysis task is improved.

The invention discloses a robot environment identification method based on multi-modal fusion, and the method comprises the following steps: collecting image data, point cloud data and acoustic data in a robot operation environment, and carrying out the preprocessing; inputting the structured multi-modal input sample into a modal perception type quantum encoder of a multi-modal quantum graph neural network; constructing a modal coupling quantum diagram based on the modal quantum state representation set; inputting the modal coupling quantum diagram into an entangled modal propagation unit, and performing cross-modal information propagation of nodes of the modal coupling quantum diagram through a parameterized quantum circuit; quantum measurement operation is executed on the fusion quantum graph embedded representation, and an environment recognition result is output; and inputting an environment identification result into a robot control module, and driving the robot to execute path adjustment, obstacle avoidance or action response. According to the method, the multi-mode quantum graph neural network is adopted, and autonomous recognition of the robot in a complex environment is achieved.

The invention provides a wireless communication method based on a cellular Internet of Vehicles technology, a communication terminal and a vehicle, and relates to the technical field of vehicle communication. The method comprises the following steps: adding a first vehicle-mounted communication terminal and a second vehicle-mounted communication terminal of a target following vehicle in preset formation vehicles to a communication network of a target multicast group, and generating a target multicast routing table, so that a leading vehicle adopts a multicast communication mode, the multicast data can be sent to the second vehicle-mounted communication terminal of each target following vehicle in a targeted manner through the first vehicle-mounted communication terminal, so that the data is prevented from being sent to other vehicles. Due to the adoption of the multicast communication mode, the vehicle-mounted communication terminal of any vehicle in the target multicast group sends the multicast data, and the vehicle-mounted communication terminals of the other vehicles can receive the multicast data, so that unnecessary repeated transmission is avoided, the risk of network congestion is remarkably reduced, the information spreading efficiency is improved, and the user experience is improved. And meanwhile, all target following vehicles can receive the latest instruction in time, and synchronization and stability of a motorcade are maintained.

1. The name of the design product: floor advertising machine (32 inches). 2. The use of the design product: the design product is used for advertising display and information dissemination. 3. The design points of the design product: in shape. 4. The picture or photo that best indicates the design points: perspective view 1.

The invention discloses a shared power bank battery safety early warning method and system based on big data, and relates to the technical field of battery safety management, and the method comprises the steps: constructing an information flow network, extracting the information propagation intensity of a directed edge, calculating the abnormal propagation rate of equipment, and generating a cumulative propagation effect in a time window. And calculating a fractional derivative of the cumulative propagation effect, constructing an adjacent matrix and calculating a maximum characteristic value, and calculating the spectral diffusion contribution and a comprehensive abnormal value of the equipment in combination with the cumulative propagation effect and the maximum characteristic value. According to the method, the dynamic strain field is combined with the flow field velocity analysis, the spatial distribution description capability of the internal structural risk of the battery is enhanced, the comprehensive coupling factor is combined with the information flow network propagation model, and the abnormal propagation mechanism modeling capability of the shared power bank in the large-scale mass use process is improved.

This application discloses a method for sharing a customized in-vehicle system function scheme, comprising: receiving a sharing request from an electronic device for a customized in-vehicle system function scheme that has been edited; querying the editing information of the customized scheme based on the scene identification code in the sharing request; encrypting the editing information to generate encrypted editing information; generating a sharing link carrying the encrypted editing information; and sending the sharing link to the electronic device, so that the customized scheme can be viewed and inherited by the recipient after sharing. The method for sharing a customized in-vehicle system function scheme of this application can disseminate the content and editing information of the customized scheme in the form of a sharing link through an encryption process, so that the recipient can view and inherit the customized scheme. This allows the content of the customized scheme and the editing information of the sharing party to be disseminated, reducing user operations and reducing the risk of information leakage during dissemination to a certain extent, making the information dissemination process more secure.

The invention discloses an unmanned aerial vehicle target dynamic locking method based on multi-source sensing fusion. The method comprises the following steps of 1, collecting and standardizing multi-source original data from an unmanned aerial vehicle sensor; 2, performing feature extraction on the standardized data set to obtain a data feature vector; 3, calculating the similarity between the data feature vectors by using a Manhattan distance, and establishing a Manhattan graph; 4, performing causal analysis on the Manhattan graph to construct a causal relationship graph; 5, performing K-means clustering on nodes in the causal relationship graph to obtain an optimized causal relationship graph; step 6, calculating a shortest propagation path in the optimized causal diagram by using an improved BellmanFord algorithm, and identifying key nodes to obtain an optimal information propagation path; and 7, determining the current predicted position state of the target, and executing the target locking action of the unmanned aerial vehicle. According to the method, causal analysis and an improved BellmanFord algorithm are fused, and the target of the unmanned aerial vehicle is dynamically and accurately locked.

The present application relates to a kind of based on dynamic spectrum diagram topology infringement counterfeit commodity intelligent identification system, belong to image visual identification technical field.System includes data acquisition, pre-processing, time sequence feature extraction, pixel correlation diagram construction, graph topology feature calculation, infringement counterfeit identification etc.Module.Data acquisition module generates dynamic spectrum sequence containing coordinate pixel etc.Information, time sequence feature extraction module extracts attenuation half-life etc.Feature, pixel correlation diagram construction module takes pixel as node, time sequence feature as node feature and based on DTW distance calculation edge weight, graph topology feature calculation module obtains node connection close and information propagation feature.This system discards traditional commodity appearance comparison mode, utilizes commodity physical characteristic time sequence law identification, avoids high imitation product visual deception, significantly improves identification accuracy.

The invention discloses a representation vector optimization method, system and device for sequence data prediction and a medium, and particularly relates to the technical field of representation vector optimization. According to the method, a hidden variable is introduced into each variable, meanwhile, hidden state coding is carried out on original observation variables full of uncertainty, probability distribution obeyed by the hidden variables is estimated by using a neural network, and parameters of the distribution serve as more robust representation vectors of the variables; the method comprises the following steps: adaptively screening out an invariant feature code which is most stable in association with a predicted target from a variable sequence by using a gating mechanism so as to reduce the uncertainty of the invariant feature code; meanwhile, through an asynchronous loop optimization strategy, information gains provided by the node variables and the information variables in the information spreading process are continuously obtained, and representation information of the node variables and the information variables under information coupling is enhanced.

The invention discloses a sea condition classification method and system based on a double-branch graph enhanced space-time network, and relates to the technical field of computers, and the method comprises the steps: receiving a multi-channel ship motion time sequence, carrying out the down-sampling of the multi-channel ship motion time sequence into an interlaced subsequence, and carrying out the down-sampling of the interlaced subsequence; performing feature extraction on each staggered subsequence based on a multi-layer perceptron, outputting compressed feature representation, and splicing all the compressed feature representations of the staggered subsequences to form time compression features; dynamically constructing an adjacent matrix between channels based on the time compression feature, and executing a graph convolution operation for information propagation to obtain a graph enhancement feature; and performing two-way fusion and classification based on the graph enhancement feature and the time compression feature, and outputting a sea condition classification result. According to the invention, through an efficient and lightweight network model, rapid classification of sea condition levels is completed on a resource-limited embedded device.

The invention belongs to the technical field of artificial intelligence and energy time sequence prediction, and provides an energy time sequence prediction method and device based on frequency domain decoupling and multi-scale fusion. The method comprises the following steps: 1) carrying out feature embedding and normalization preprocessing on input energy time sequence data; 2) carrying out frequency domain analysis by adopting fast Fourier transform, and carrying out multi-scale division and reconstruction according to the period length of the main periodic component to extract multi-scale periodic features; 3) constructing a double-branch structure combining a gating unit and a multi-head attention mechanism, and respectively modeling long-term trend dependence and short-term dynamic characteristics; 4) distinguishing a homogeneity and heterogeneity relationship between variables based on a Pearson's correlation coefficient; 5) utilizing the graph convolutional network to realize structured information propagation among multiple variables, and outputting prediction results of a plurality of time steps in the future through a linear projection layer; and 6) deploying the model in various energy monitoring systems for long-time-sequence prediction of energy scenes such as natural gas flow, power load and solar energy yield. According to the method, the incidence relation between the multi-scale period rule and the complex variable of the energy time sequence data can be effectively captured, so that the prediction precision of an energy system and the intelligent level of operation scheduling are improved.

A silicon-anchored hardware enforcement system enables low-latency, cryptographically gated execution of artificial intelligence for stem cell therapy and regenerative medicine. A Sovereign Identity Token derived from a Physical Unclonable Function (PUF) permanently binds model decryption to a specific hardware instance and restricts execution to a Trusted Execution Environment synchronized to a hardware-protected Safety Epoch. Encrypted AI model weights are decryptable only upon successful hardware validation. Multimodal clinical inputs, including genetic, imaging, and structured patient data, are integrated through a hardware-constrained fusion architecture that operates under enforced biological safety parameters. FPGA-implemented predicate logic evaluates defined biological safety conditions at sub-millisecond latency, and an ASIC-based nullification circuit irreversibly suppresses outputs that violate hardware-defined thresholds within a bounded millisecond response time. A distributed revocation protocol propagates credential invalidation across networked nodes within sub-second latency while preserving reduced-capacity safe mode operation. A permissioned provenance ledger records hardware-attested execution events and supports automated regulatory documentation. The system provides secure clinical deployment, federated research enablement, and verifiable auditability for high-risk therapeutic environments.

The method provides for one or more processors to disperse object data for storage within a fifth-generation radio access network (RAN). The one or more processors receive radio frequency (RF) input for object data storage from a client device. The one or more processors perform a setup session configuring the RF input received for object data storage. The one or more processors perform a translation of the received RF input, wherein the translation enables processing of the RF input by an information dispersal algorithm (IDA) and enables storage of the object data of the RF input among multiple next generation node base stations (gNBs) forming a gNB cluster within a radio access network (RAN), and the one or more processors storing the object data of the received RF input across the gNB cluster in an Object Segment format.

The application provides an information processing method and server based on unified code issuing, and relates to an encoding medium related to digital currency circulation or digital information transfer. The application can realize the value quantification and redemption of digital currency based on the circulation field by using the information processing technology based on unified code issuing, various sensing access modes, scene-based specific theme access, and accurate tracing of each node in the information propagation process.

This invention discloses an adaptive graph clustering method based on feature adversarial and graph Transformer, belonging to the field of image processing technology. The original graph is optimized by a feature adversarial autoencoder module to match prior distributions of representation features, obtaining attribute-level features. The feature adversarial autoencoder module includes an autoencoder and a feature adversarial module. The original graph is then processed by a graph Transformer autoencoder module to perform the convergence and fusion of local features and global structure, and information propagation, obtaining structural-level features. Adaptive fusion of attribute-level and structural-level dual-source features from the feature adversarial autoencoder module and the graph Transformer autoencoder module completes graph clustering. The training of the feature adversarial autoencoder module and the graph Transformer autoencoder module employs a self-supervised learning approach. A joint optimization loss function guides the joint optimization of graph representation learning and cluster assignment. The joint optimization loss function includes a feature adversarial loss function, an optimization loss function, and a self-supervised learning loss function. The feature adversarial loss function includes the feature reconstruction loss function of the autoencoder and the minimum cross-entropy loss optimization function of the feature adversarial module.

The present invention provides a computer program, an information processing method, and an information processing device that predict temperature changes in an image output device in advance and perform appropriate thermal control. [Solution] The information processing device includes an input layer that inputs a time-series first thermal information based on an image input to an image output device; a thermal reservoir layer having a plurality of virtual nodes that hold a time-series second thermal information observable at each of a plurality of locations as the first thermal information propagates to a plurality of locations of the image output device based on the characteristics including the thermal characteristics of the image output device; a coupling layer that linearly combines the time-series second thermal information held by the plurality of virtual nodes of the thermal reservoir layer using coupling coefficients between the plurality of virtual nodes based on the time-series first thermal information; and an output layer that outputs the linearly combined second thermal information as an output value; and a determination unit that determines the coupling coefficients based on training data including the first thermal information based on the image and the correct values ​​of the output value.

The invention discloses a road surface appearance state pixel-level intelligent detection system based on continuous coding and decoding, relates to the technical field of road intelligent operation and maintenance, and solves the problems of insufficient road surface detection precision and generalization ability caused by incomplete coding-decoding process and unreasonable information equivalent processing of a current semantic segmentation network. According to the invention, a multi-layer information interaction module is connected with adjacent stages of a multi-stage multi-layer continuous coding module and a multi-stage multi-layer continuous decoding module; the information spreading unit transmits the feature extraction result of each layer of the previous stage in the multi-stage multi-layer continuous coding module and the multi-layer continuous decoding module to the feature extraction result of the same layer of the next stage; the learnable feature reconstruction channel segmentation marking module is connected with the multi-stage multi-layer continuous coding module and the multi-stage multi-layer continuous decoding module; according to the invention, after training is carried out based on a large-scale asphalt pavement image data set, detection of various types of pavement apparent diseases and design elements with pixel-level precision can be synchronously realized.

Embodiments of the present disclosure provide a solution for video processing. A method for video processing is proposed. The method comprises: determining, for a conversion between a current video unit within a current frame of a video and a bitstream of the video, a scaling factor based on coding information of the current frame; adjusting a strength factor for the current video unit based on the scaling factor; determining a quantization parameter (QP) for the current video unit based on the adjusted strength factor and information propagation associated with the current video unit; and performing the conversion based on the QP.

The invention provides an e-commerce commodity category adaptive recognition and semantic evaluation method, which comprises the following steps of: extracting path hops and associated edge semantic type labels between target commodities from a commodity relation graph, and extracting all shortest path sets through graph traversal to obtain a hop distance between the commodities and an edge type combination; performing multi-layer information propagation on the hop count distance between the commodities by adopting a multi-layer graph convolutional neural network through the information transmission loss value, and determining a commodity semantic embedding vector after propagation of each layer; according to the propagated commodity semantic embedding vector of each layer, the vector distance between commodity semantic embedding is extracted, the similarity is evaluated, and the optimized information propagation aggregation layer number is obtained based on the similarity convergence trend; and evaluating the coincidence degree of the commodity category inference result and the actual commodity category attribution, and identifying the category inference deviation to obtain the correction direction of the self-adaptive identification of the e-commerce commodity category.

The invention relates to a coal mine underground gas inspection collaborative decision-making method and a coal mine underground gas inspection collaborative decision-making device. Comprising the steps of amphibious inspection robot system initialization, inertial data preprocessing and state prediction, point cloud acquisition and state updating, multi-source data fusion and map updating, spatio-temporal feature node construction, dynamic spatio-temporal diagram modeling and information spreading, emission source probability field generation, three-dimensional semantic understanding, terrain geometric parameter extraction and trafficability quantitative decision making. The gas emission source positioning method has the advantage that the gas emission source positioning method based on dynamic space-time diagram modeling and physical information fusion is provided. A node network with space-time constraint is constructed, a learnable neural network is utilized to calculate node association weights, and a gating modulation mechanism of environmental factors such as wind speed is combined, so that efficient fusion and reasoning of sparse and dynamic gas data are realized. An attention mechanism and kernel density estimation capable of optimizing bandwidth are further introduced, a continuous and explainable gas emission probability field is generated, and the limitation of traditional point type monitoring is broken through.

The invention relates to the technical field of fire safety monitoring, and discloses a fire-fighting equipment fault prediction method based on deep learning. According to the method, multi-source operation data of a fire fighting system and an alarm / linkage system are collected, an equipment topological graph is constructed, and a two-channel neural network model fusing time sequence convolution and graph convolution is established. Adaptivity and information propagation efficiency of a topological structure are enhanced through spectrum gap optimization and a greedy topology reconnection mechanism, and flexible alignment and stable fusion of time sequence features and structural features are realized by introducing a multi-modal mixed comparison consistency fusion mechanism. The method can identify the potential fault of the fire-fighting equipment in advance under a complex working condition, realizes accurate prediction and graded alarm, improves the reliability, robustness and intelligent operation and maintenance capability of the system, and is suitable for the field of operation safety monitoring of building fire-fighting systems and industrial fire-fighting equipment.

This invention discloses a method, device, and medium for fusing network active detection and computing power perception, belonging to the field of computing power network and mobile communication technology. The technical problem this invention aims to solve is how to obtain computing power information consistent with network information during computing network information perception, thereby improving the consistency and efficiency of computing network information dissemination. The technical solution adopted is as follows: In a computing power network scenario, computing power information is carried through protocol extensions in network active detection requests and responses: the entry node sends a computing power query request to the exit node, and the exit node replies with the computing power query result to the entry node; the exit node maintains an edge computing power information database, and the computing power information requested by the entry node is searched in the edge computing power information database; when the computing power service information changes and the change exceeds a set threshold, the exit node is notified; the exit node updates the edge computing power information database and notifies the entry node, through independent messages or network perception messages.

The CMOS neuron-synaptic unit circuit system suitable for the spiking neural network is realized, so that network parameters of the spiking neural network are effectively initialized, and the weight is dynamically adjusted to promote information propagation and learning. The CMOS neuron-synaptic unit circuit system comprises a front LIF neuron circuit module which receives an input pulse signal and generates an output pulse through integration; the synapse circuit receives the output pulse and adjusts the transmission intensity of the signal by adjusting the real-time weight; the STDP circuit adjusts the synaptic weight according to the time difference between the pulse output by the front LIF neuron circuit module and the pulse output by the rear LIF neuron circuit module; the ATML circuit receives the output pulse of the front LIF neuron circuit module and the output pulse of the rear LIF neuron circuit module, and adjusts the weight of the STDP to update the length of a time window by changing the size of a time window length signal Vleak; and the LIF neuron circuit receives the signal from the synaptic circuit module, and simulates the neuron signal integration and pulse generation process again.

This invention discloses a dynamic clustering method for space-based distributed collaborative detection. First, based on dynamically changing mission requirements and target status, a dynamic clustering algorithm optimized for collaborative detection performance is designed to cluster satellites within the constellation. Then, using the comprehensive capabilities of nodes as a crucial basis for node star election and cluster structure maintenance, a multi-factor weighted dynamic edge node star selection algorithm is designed, ultimately achieving a hierarchical, ordered, distributed collaborative architecture of "constellation-cluster-single star". This invention can reduce the information propagation cost within the constellation and improve information collaboration and mission execution efficiency.

The application discloses a method and device for determining node influence, relates to the technical field of network analysis, and is used for improving the accuracy of determining user influence, so as to improve the efficiency of resource allocation and the efficiency of information transmission in a network. The method comprises the following steps: acquiring access information of a plurality of nodes in the network; wherein the access information is information of accessing the network; determining a second node adjacent to a first node and a third node connected to the first node in the network according to the similarity between the access information of the plurality of nodes; wherein the plurality of nodes comprise the first node, the second node and the third node; determining local influence of the first node according to the similarity between the access information of the first node and the access information of the second node; determining global influence of the first node according to the similarity between the access information of the first node and the access information of the third node; and determining influence of the first node in the network according to the local influence and the global influence.

The application provides a graph adversarial generation and dynamic graph representation learning method under unbalanced medical data, belongs to the technical field of medical data analysis and the technical field of machine learning, and the method comprises the following steps: S1. data processing and generation; S2. data set balancing; and S3. severe prediction model construction. The application mainly solves two main problems: (1) the unbalance of data, which seriously leads to the model tending to the majority class and influences the prediction stability; and (2) a real data set is usually complex and high-dimensional, and it is difficult to sufficiently learn between nodes, so that information transmission is limited in local neighbors, and the prediction accuracy is reduced.

The invention relates to the technical field of data analysis, in particular to a news propagation influence quantitative evaluation method and system based on data mining, and the system comprises a data collection and preprocessing module, a dynamic index management module, an intelligent modeling and analysis module, a causal inference and explanation module and a storage and retrieval module. In order to solve the problem that a traditional evaluation model is insufficient in scene adaptability due to the fact that a fixed weight and a linear attenuation hypothesis are adopted, a dynamic attenuation model and a reinforcement learning mechanism are introduced in the scheme, and chain reactions in the information spreading process are described through modeling of a self-excitation effect in a spreading event; on the basis, an evaluation index weight distribution mechanism is dynamically optimized in combination with an adaptive algorithm, so that the model can automatically adapt to rapid outbreak of sudden news and long-acting propagation characteristics of policy news, and a nonlinear rule presented in the propagation process is accurately captured; according to the method, the scene adaptability and timeliness of the evaluation result are effectively improved.