23 results about "Network mapping" patented technology

The invention relates to a method (100) for configuring a mesh topology Ethernet optical network connecting a plurality of devices and switches. Each switch includes at least one transceiver. Each transceiver has a transmitting section and a receiving section. Each transceiver section is associated with a virtual port. The method (100) includes an initialization (101), a network mapping (103) including the identification of possible links enabling communication between two devices, a possible link being a sequence of virtual ports enabling communication. The method (100) then includes a definition (104) of a network configuration including selected links, a deployment (105) of the network configuration including the disabling of virtual ports not used by the selected links, and a power-off (106) of the transceiver sections equipping each disabled virtual port.Figure for the abbreviation: [Fig. 1].

A model that includes a generative network mapping a latent vector to a feature vector, wherein weights in the generative network are modelled as probabilistic distributions, is trained on observed data points with incomplete observations of the feature vector to learn values of the weights. From a plurality of potential next features to observe, a target feature is selected that maximizes a measure of expected reduction in uncertainty in the distribution of the weights given the observed data points so far. A target data point comprising at least the target feature is then requested, and the generative network is further trained based on the target data point received in response to the request.

The application provides an arbitrary style transfer enhancement method, a computer device and a program product. The method constructs a content-style-style triple data set, filters candidate samples according to content fidelity, style consistency and aesthetic score, and obtains high-quality triples. Based on the data set, an attention block denoising diffusion model is constructed, a frozen variational autoencoder is used to encode an image into a latent space token, and a noisy latent token is generated for denoising generation. In the training stage, the supervision fine-tuning and online reinforcement learning training of the denoising loss are combined, content rewards and style rewards are introduced, the rewards are embedded through a learnable projection network and injected into the attention block to modulate the denoising process, thereby suppressing artifacts and improving content fidelity and style consistency. In the inference stage, the input is a to-be-styled image and a style reference image, and the output is a target styled image. In the case of fewer trainable parameters, single inference output is supported, and the reference projection is cached to reduce repeated calculation.

This disclosure presents an intelligent star map recognition method, apparatus, device, and medium based on multi-network mapping. One specific implementation of the method includes: extracting features from a star image using a pre-constructed star feature extraction network and an extracted set of star information to generate feature information for each node; constructing a graph network from the star information set based on star-sensor camera parameters to generate a star network structure; performing star map recognition on the star network structure using a pre-constructed star map recognition network model, the feature information for each node, and prior attitude information to generate a star recognition result; caching the star recognition result; and outputting the star recognition result to the spacecraft's attitude control system in real time. This implementation can ensure a high star recognition rate while significantly reducing the storage complexity of the feature library and the computational complexity of the algorithm, and exhibits stronger robustness to changes in the number of stars and noise in star magnitude information.

The application provides a multi-stage semantic emotion-driven digital human action generation method, and relates to the technical field of digital human action generation. The application extracts audio features and text semantic features and completes time alignment and fusion through multi-modal preprocessing of input audio and text to obtain a multi-modal feature sequence. Then, the sequence is used to identify semantic intent and analyze emotion to obtain a semantic intent sequence and an emotion curve containing categories, intensity and change trend. Subsequently, a global action script containing action type, intensity, duration and priority is generated according to a preset action mapping rule. The script is encoded into a feature vector and input into a lightweight neural network to obtain an action parameter sequence containing facial expressions, mouth shapes, skeletal actions and gaze directions. After time sequence smoothing and physical constraint checking, the final digital human driving parameters are output, which can realize low-power real-time operation, smooth action, accurate mouth shape synchronization and emotion and semantic matching.

Techniques are disclosed for mapping an indoor space using light fidelity (Li-Fi) communication. An entity may obtain location information for a plurality of Li-Fi devices, that includes at least one mobile Li-Fi device. The entity may determine, based on the location information, a mapping of the indoor space, where the mapping comprises spatial information for one or more continuous areas of the indoor space based on LOS Li-Fi communication between Li-Fi devices of the plurality of Li-Fi devices. The entity may be a Li-Fi device or a server. The entity may assign locations of Li-Fi devices in LOS communication with each other to the same continuous area to help determine the mapping.

Systems, devices, and methods support network traffic, for example, Broadcast, unknown Unicast, and link-local Multicast (BUM) traffic, in an overlay by utilizing Protocol Independent Multicast (PIM)-Source-Specific Multicast (SSM) in an underlay, while removing dependency on PIM-Any Source Multicast (ASM). A mapping system receives a mapping registration message and / or a mapping request message of a candidate device configured with a virtual network instance, for example, a Layer 2 Virtual Network Instance (L2VNI) implemented with layer 2 flooding. In response to receiving the mapping registration message, the mapping system transmits a mapping notification message to one or more member devices of an underlay group associated with the L2VNI. The mapping notification message indicates that the candidate device has joined the underlay group. In response to receiving the mapping request message, the mapping system transmits, to the candidate device, a list indicating that the member device(s) intends to transmit the network traffic.

This invention provides a method and system for aggregating multi-source heterogeneous intelligence data. The aggregation method includes: acquiring raw multi-source heterogeneous scientific and technological intelligence data; extracting structured entropy representing structural complexity and source domain fingerprint representing source reliability; storing all three in a data lake for unstructured storage; receiving a query request; parsing the query entity and relational pattern; initially screening candidate data units by combining structured entropy and source domain fingerprint; performing entity identification on the candidate data and linking it to a pre-set scientific and technological knowledge graph; constructing a confidence model through graph topology and semantic information to complete entity referencing resolution; extracting semantic vectors from multimodal data instances of the same knowledge graph entity using the multimodal model; mapping them to a shared Hamming space via a cross-modal hash network; aligning cross-source facts based on Hamming distance; retrieving isomorphic subgraphs in the knowledge graph as aggregation patterns based on query relational patterns; and integrating the aligned data to generate structured scientific and technological intelligence.

This invention discloses a satellite network digital twin simulation system and method based on state coordination, primarily addressing the problem that existing technologies cannot collaboratively simulate future network states, leading to poor adaptability in routing decisions. The system comprises a satellite network entity layer, a twin data resource layer, a virtual twin modeling layer, a simulation and analysis layer, and an application service execution layer. The simulation and analysis layer includes a satellite-ground collaborative simulation module, comprising a GAT-TCN link load prediction model and an ST-Transformer ground traffic prediction model. These models are used to perform time-series predictions of inter-satellite link load and satellite-ground access traffic, respectively, forming a global network mapping and enabling collaborative evolution simulation of satellite and ground states. This allows routing strategies to be formulated based on future network evolution trends. This invention can accurately reproduce future physical network operating scenarios, improve the adaptability of routing decisions, and can be used to simulate future network states in large-scale satellite networks.

The utility model discloses a kind of acoustic positioning installation equipment based on acoustic wave type gas pipe network surveying and mapping positioning, specifically related to gas pipe positioning technical field, including arc frame, the inner wall of arc frame is fixedly connected in guide rail, the outer wall of guide rail is equipped with guiding positioning installation component, guiding positioning installation component includes: arc sleeve, slidingly connected on the outer wall of guide rail, arc sleeve moves along the circular arc path of arc frame inner wall;Ultrasonic distance sensor, fixedly located in one side of arc sleeve, and another side of arc sleeve is fixedly installed with holding rod.The utility model uses guiding positioning installation component, by hand holding on the outer wall of holding rod to carry out arc-shaped up, arc sleeve carries out arc-shaped path up along the inner wall of arc frame, ultrasonic distance sensor carries out ultrasonic distance sensing along the arc-shaped path of gas pipeline outside, can realize the positioning installation of large-area arc-shaped path.

This invention belongs to the technical field of security auditing in computer data processing, and relates to a database auditing method and system based on intelligent translation and three-layer association. The method includes the following steps: synthesizing an authorization seed tag and writing it into the system kernel shared memory area; extracting network mapping instruction clusters and native control command byte sequences containing underlying communication source port parameters corresponding to random source port identifiers; stripping port information to generate a unique legitimate addressing index upon successful verification; extracting fine-grained instruction slice signals; constructing memory coordinate offset vectors indicating specific harmful action behavior characteristics; continuously mirroring and copying the corresponding human natural language bit phrase sequences; and encapsulating and outputting a high-aggregation semantic warning data packet. This invention addresses the vulnerabilities in the association between terminal identity and database operations, the insufficient intuitive readability of operational behaviors, and the need to improve the security of cross-layer information fusion in multi-layer architectures in traditional database auditing methods.

The application provides a chip end face image definition evaluation method and system based on semantic information, comprising: obtaining a chip end face image to be evaluated; inputting the chip end face image to an image degradation feature extraction module, extracting the degradation features of the image and performing global average pooling processing, and obtaining a guide vector through MLP network mapping; constructing an adaptive prompt word pair, processing the image to be evaluated through an image encoder to obtain an image feature vector, and simultaneously processing the adaptive prompt word pair through a text encoder to obtain a prompt word feature vector; calculating the definition score of the image to be evaluated through a Softmax function and comparing it with a preset threshold to obtain a determination result. The application overcomes the problem that the picture definition evaluation scale is not the same under the condition that the texture quantity and density are different, and makes the evaluation score only depend on the overall definition of the picture, so as to avoid the problem that the definition of the defects at different focal depths of the chip end face image causes the score to be biased and leads to inaccuracy.

A method for constructing a message routing table for heterogeneous protocol communication comprises: (a) when at least one service discovery message is input to an i-th routing node—the i-th routing node is included in an i-th system, and the i-th system is one of a first system to an n-th system—a computing device corresponding to the i-th system causes the i-th routing node to generate i-th service map information by (a_i) analyzing the service discovery message through an i-th service discovery module included in the i-th routing node; and (a_ii) enabling the i-th network information and i-th resource information of the i-th routing node and other network information and other resource information of at least one other routing node to be shared with each other through an i-th network bus manager module included in the i-th routing node, and generating i-th network map information and i-th resource map information using the i-th network information and i-th resource information and the other network information and other resource information shared from the at least one other routing node; and (b) a step of constructing an i-th distributed routing table by causing the computing device to cause the i-th routing node to analyze the i-th service map information, the i-th network map information, and the i-th resource map information through an i-th routing table builder module included in the i-th routing node; the method is provided.

The application discloses a big data intelligent financial customer acquisition system and relates to the technical field of financial technology and big data technology. A data acquisition module is used for acquiring data, processing the data through standardization, generating an internal element set and a risk event set; a graph construction module is used for constructing an industrial correlation knowledge graph from data acquired from a global enterprise database; a conduction network module is used for calculating and marking a conduction strength coefficient and an attenuation factor, and constructing a dynamic conduction network; a mapping set module is used for identifying a risk conduction entry node and forming a mapping set; a walk simulation module is used for calculating a risk conduction strength and a probability through Monte Carlo random walk simulation, and forming a risk conduction strength distribution graph; and a customer activation module is used for dynamically reducing an original activation value score of a customer entity to generate a comprehensive value sequence, and generating a customer activation strategy. The application effectively solves the problems of insufficient risk foresight and low marketing efficiency of the existing system.

This invention provides a method and system for simulating rock creep processes based on deep learning-based near-field dynamics, involving the interdisciplinary fields of rock mechanics, computational mechanics, and artificial intelligence. The method includes: acquiring parameter information of the target rock to be simulated, including the rock's geometry and physical properties; discretizing the target rock into several material points based on the parameter information to construct a creep damage model; performing time-step iterative simulation of the rock creep process based on the initialized creep damage model, ultimately obtaining the displacement field, stress field, and creep strain distribution of the target rock over time. This invention significantly improves computational efficiency by establishing a neural network mapping from bond geometric variables to force density, replacing the complex iterative calculations in traditional near-field dynamics. Furthermore, by combining the creep damage model with damage evolution criteria, it achieves high-precision simulation of the entire creep process of fractured rock masses.

The application discloses a numerical simulation method for estimating equivalent seepage of a three-dimensional fracture network based on topological properties, which is based on relevant data of a fracture network of a research area, first constructs a small three-dimensional fracture network sample set, then maps the sample fracture network into an equivalent two-dimensional undirected graph by using graph theory, solves equivalent permeability, and constitutes a training sample set; then the maximum information index method is used to mine main control factors of the equivalent permeability of the fracture network, and the control relationship between the fracture network parameters and the underground flow field is clarified; finally, a substitution model between the main control factors and the equivalent permeability is constructed based on a BPNN machine learning model, and conversion of a target discrete fracture network to an equivalent medium is realized. The application provides a simulation scheme with high prediction accuracy and low calculation cost reflecting topological connectivity for numerical simulation of an underground flow field of a fractured stratum at a site scale.

The application discloses a relay core selection method and system for a communication distance limited connection, belongs to the field of pulse neural network compilation, and aims at the problem that the direct communication distance between cores in a neural quasi-state computer is limited, the relay resource capacity and availability are limited, and long-distance connection needs to be assisted by a relay core. In the application, connection demand and core state are acquired in a compilation stage, a network flow graph with capacity and distance constraints is constructed, a minimum cost flow is solved, and flow decomposition is performed on the result to obtain a relay path and a rule configuration scheme. A relay rule set that can be written into hardware and storage usage of each relay core are outputted, and are used for relay planning of pulse neural network mapping deployment, so that the ultra-distance connection reachability is ensured and the relay rule capacity utilization efficiency is improved in an actual hardware system. The application has strong universality, and can be adapted to neural quasi-state chips with different topological structures, distance measurement modes and maximum communication distance parameters.

The application relates to the technical field of power equipment intelligent engineering, and discloses a digital collaborative optimization design method for structure parameters of an extra-high voltage insulator, which comprises the following steps: discretizing an initial model to generate a joint arrangement field containing a non-Euclidean graph structure and a joint attribute vector; mapping surface features and calculating dynamic connection weights by using a graph attention network; inputting a physical information graph neural network, combining physical equation residual error training, and reasoning a multi-field physical state; defining a topological interference penalty function according to a manufacturing envelope abstract scalar flow field; constructing a reinforcement learning intelligent agent, taking a graph structure and a multi-field state as a state space, taking node topology, geometry and material updating as an action space, and relying on a comprehensive reward function to drive interaction to output an updated result; and decoding, reconstructing and executing a multi-physical field simulation check. The application avoids parameter conflicts caused by hierarchical independent deduction, and realizes multi-dimensional parameter joint calculation under multi-field coupling conditions.

This invention provides a method for attitude maintenance and trajectory tracking control of an unmanned surface vessel under complex sea conditions, belonging to the field of ship automatic control and intelligent navigation technology. It establishes a three-degree-of-freedom planar motion model of the unmanned surface vessel by real-time data acquisition and synchronous preprocessing from multiple sensors, and obtains stable attitude and velocity states through extended state fusion estimation. Under a unified time base, wave feature parameters are extracted through sliding window frequency domain analysis, and disturbance intensity indices are constructed to achieve online sea state discrimination. The tracking error is calculated based on trajectory and state estimation, and control coefficients are generated through network mapping. Performance indices are reconstructed to obtain baseline control components, which are then superimposed with fast compensation terms. Multiple state indices are concatenated into a state vector, input into an anti-disturbance network to generate dynamic feedforward compensation, and then fused with the compensation control before issuing execution commands. Periodic looping achieves closed-loop adaptive control. This method realizes coordinated control of attitude stabilization and high-precision trajectory tracking under complex sea conditions.

The application discloses a machine learning-based digital twin dynamic optimization method, device and medium, relates to the technical field of network security, and solves the problem that a to-be-optimized node in a network cannot be accurately screened based on an influence degree. The method comprises the following steps: acquiring network structure data and historical abnormal data of a target network, and mapping the target network to a digital network space according to the network structure data and the historical abnormal data; analyzing different transmission nodes in the digital network space; analyzing different transmission nodes in the digital network space according to the historical abnormal data, to obtain a first transmission node and a second transmission node; analyzing the first transmission node and the second transmission node, to obtain a to-be-optimized node in the digital network space; and optimizing a corresponding network node in the target network according to the to-be-optimized node in the digital network space. The application realizes accurate screening of a to-be-optimized node in a network based on an influence degree.

The application provides a new energy industry data deep mining and visualization method and equipment based on Pandas, and relates to the technical field of new energy industry data mining and analysis. The method first uses a network mapping platform for deep data mining, imports a relational database, then uses Pandas to clean and secondarily develop the industry data to obtain high-quality data, constructs an industry database meeting the depth of new energy industry demand, and supports on-demand retrieval and data visualization. The application can accurately locate specific data assets in the new energy field by constructing an aggregation mapping search model based on industry basic words and sensitive words. This has significance for the new energy industry in that it can strip out vertical data with real technical value and commercial value from massive Internet noise, providing a high-quality data base for accurately grasping industry dynamics and researching technical routes.

The application discloses a kind of virtual-real fusion large-scale network security simulation system and method, belong to network security simulation technical field.The system includes distributed network simulation layer, containerized service layer, virtual-real fusion bridging layer and artificial intelligence driving layer.Distributed network simulation layer is used to build logically centralized distributed network simulation environment, and the simulation of large-scale network topology and traffic forwarding behavior is realized by running network simulation instance in multiple hosts or containers;Containerized service layer is used to carry business service and security test task;Virtual-real fusion bridging layer is used to realize the network mapping and bidirectional communication between simulation node and external entity object;Artificial intelligence driving layer is used to train, inference detection model or decision model, and feed back the result to control module.The application forms the closed-loop verification of data acquisition, model processing, decision generation and environment feedback, improves the authenticity, scalability and intelligent level of network security simulation.