52 results about "Network adaptation" patented technology

The invention relates to a method and apparatus for generating a project-specific network architecture. The invention relates to a method for generating a project-specific network architecture, said method comprising the steps of:-providing (S1) a base model, in particular a large language model, with LoRa network adaptation; providing (S2) a model library having training data pairs, each of which consists of input data having a model application and / or model and / or hardware and / or software specification, and output data having at least one network architecture belonging to the respective input data; -selecting (S3) item-specific training pairs on the basis of said library of models; and-training (S4) the LoRa network of the base model on the basis of the item-specific training pair for generating the item-specific network architecture.

A 3D development engine-oriented adaptive flow loading method relates to the technical field of 3D development engines, and comprises the following steps: collecting environment state data through an environment sensing module; continuously acquiring a network state index through a network monitoring module; establishing a dynamic priority calculation model, taking the environment state data and the network state indexes as input parameters, and generating a quantitative priority score for each resource project; performing dynamic reconstruction and sorting processing on the resource loading task queue based on a priority scoring result, adjusting the execution number of concurrent loading tasks according to a real-time network state index, and selecting a corresponding resource compression strategy; and establishing a resource life cycle management system, and executing allocation optimization and recovery cleaning operation of the memory resources. According to the invention, adaptive flow loading of 3D resources can be realized, and the fluency of a large-scale scene, the network adaptability and the system resource utilization efficiency are improved.

The invention provides a resource loading method and system based on user behavior prediction, and relates to the technical field of data processing, and the method comprises the steps: collecting user behavior data; preprocessing the user behavior data; extracting data features of the preprocessed user behavior data; according to the data features, predicting resource demand weight distribution of the user through a double-layer LSTM based on fusion of an attention mechanism and interest drift perception; calculating the priority of each resource according to the prediction result of the resource demand weight distribution of the user in combination with the importance coefficient, the resource value density and the network adaptation coefficient of each resource; constructing a resource loading priority queue according to the priority of each resource from high to low; and dynamically adjusting a loading strategy of each resource in the resource loading priority queue in combination with a network state, and loading the resources as required. According to the method, the resources can be loaded as required, the resources possibly required by the user are preferentially loaded, and the resource utilization efficiency is improved.

The invention discloses a spacecraft Hall propulsion device test working medium flow stability control system, which comprises a sensing module for detecting working condition state parameters of working medium supply equipment, and an adaptation module for exploring current working condition control parameters by combining detection data of the sensing module and similar working condition control parameter empirical data fusion, the adjusting module is used for receiving the control parameters and dynamically adjusting the adjusting weight of the adjusting device according to different phase change periods, an experience library unit and a self-adaption unit are arranged in the adaption module, the experience library unit provides experience data of similar working conditions and control parameters, and the self-adaption unit provides self-adaption data of the control parameters. And the autonomous adaptation unit fuses and explores control parameters by constructing a neural network adaptation model, so that working condition automatic identification and control parameter autonomous adaptation under the environment that the working medium type or the working medium state rapidly changes are realized; the problems of low control precision and slow response speed caused by manual presetting of control parameters of different working media in a traditional control technology are solved, and stable control of the working medium flow is realized.

The invention provides a multi-scene user controllable personalized recommendation method and system based on super network adaptation. The method comprises the following steps: acquiring a natural language query input by a user; mapping the natural language query into a control condition text based on a large language model; encoding the control condition text into a condition embedding vector by using a pre-trained text encoder; inputting the condition embedding vector into a super network, and generating an adapter parameter matrix for at least one dense layer in the basic recommendation model by the super network; adding the adapter parameter matrix and an original weight matrix of a corresponding dense layer in the basic recommendation model to obtain an adapted recommendation model; and using the adapted recommendation model to predict preference scores of the user for the articles, and generating a recommendation list based on the preference scores. According to the technical scheme, the super network is introduced to dynamically generate the lightweight adapter, so that the limitation of a traditional recommendation system in the aspects of user controllability, scene adaptability and semantic comprehension is effectively solved.

The invention relates to a personalized network guarantee service pushing method and device, equipment and a storage medium. The method comprises the steps that a specific scene where a user is located is recognized through a terminal side intention perception SDK, credibility verification is carried out, and the specific scene comprises at least one of access to a preset place network, poor network quality or starting of a high-network-demand application; based on a specific scene in which the user is located, obtaining signing state data of a current service of the user; comprehensively analyzing the user service demand and the network adaptation capability based on the signing state data of the current service of the user; formulating or updating a personalized QoS (Quality of Service) network guarantee strategy matched with the specific scene, or pushing a personalized network guarantee service ordering request to a user, and performing state feedback; and executing a network service guarantee operation according to the network guarantee strategy. By adopting the method, precise scene perception and network side multi-network element cooperation can be realized, and the reach rate and the order conversion rate of the network guarantee service are improved.

The invention relates to a virtual reality and network interaction technology, and discloses a virtual space multi-person interaction synchronous control method oriented to an end-cloud collaborative architecture. Through combination of end-side prediction and cloud arbitration, low-delay transmission and global consistency of a multi-user interaction state are realized. The method specifically comprises the steps that an end side carries out prediction and feature extraction on action and voice data, and redundant transmission is reduced; the key state parameters are uploaded to a cloud end and are judged by a global consistency model and a conflict arbitration mechanism in a unified manner, and a synchronization instruction is generated; and the end side performs rendering and feedback according to a result. According to the method, end-cloud hierarchical prediction and arbitration, a priority-driven protocol and bandwidth adaptive control are combined, delay and conflicts are effectively reduced, and cooperation efficiency, immersion and network adaptability are improved. The method can be widely applied to the fields of universe, online education, remote collaboration, virtual simulation and the like.

The invention provides a multi-agent heterogeneous client-server collaborative voice interaction system and method, and relates to the technical field of voice interaction based on artificial intelligence. The voice interaction system aims to solve the problems that an existing voice interaction system is weak in safety, poor in compatibility, insufficient in weak network adaptability and the like. The system comprises a client and a server, and access security and dialogue continuity are guaranteed through multi-factor joint authentication and abnormal dialogue compensation; the intelligent agent capability is split into independent modules, and unified adaptation and dynamic management of the multi-source heterogeneous intelligent agents are achieved through unique identification; and through a client local cache and module decoupling fault-tolerant mechanism, the system stability in weak network and fault scenes is improved. According to the invention, agent capability abstraction, resource decoupling and end cloud collaboration are realized, a voice interaction architecture with high expansibility, high stability and strong adaptability is constructed, and the method is suitable for various application scenes such as SDK integration and multi-role accompanying chat.

The invention discloses an LWIP TTE network adaptation system based on a static ARP table. The LWIP TTE network adaptation system comprises a transmission module, a protocol conversion module, an LWIP protocol stack and a network calling module. The transmission module enables the standard TTE frame to be normally transmitted on the TTE switch; the protocol conversion module is used for protocol adaptation and data conversion; the LWIP protocol stack is used for packaging data needing to be sent and analyzing a received standard IP data packet; and the network calling module is used for transmitting the data and the target IP address information to the LWIP protocol stack and receiving a data packet fed back by the LWIP protocol stack. Protocol adaptation and data conversion are carried out through the protocol conversion module, the communication barrier between an IP protocol and a TTE protocol is broken, and conversion from an IP address to a communication identifier supported by TTE is realized; the network adaptation system uses the simple query logic of the LWIP protocol stack and the static address mapping table, so that the embedded program developed by depending on the TCP / IP protocol stack can be accessed to the TTE network at low cost, and the complexity of program migration work is simplified.

The invention provides a network adaptation method and device of a built-in proxy browser, equipment and a medium. The method comprises the steps that the built-in proxy browser comprises a relay acceleration link and a direct connection link; a relay acceleration link is used by default; setting a timer, regularly accessing a set website through a local client, and judging network connectivity through an HTTPS request; if the HTTPS requests to detect that the network is connected, judging whether the link used by the current built-in proxy browser is a direct connection link or not, and if so, not changing; if not, using a direct connection link; if the HTTPS request detects that the network is not connected, judging whether the current built-in proxy server uses a relay acceleration link or not, and if so, not changing; if not, the built-in proxy browser uses a relay acceleration link for connection; therefore, the browser can switch the line in time, the network access speed is ensured, and the operation cost of an enterprise can be reduced.

The invention discloses a federal compression learning method and system based on knowledge distillation, and the method comprises the steps: enabling each terminal device to adaptively adjust the compression ratio of the terminal device according to a real-time network, and carrying out the compression sampling through the adjusted compression ratio; then, taking a compressed sampling value as a knowledge distillation medium between edge equipment and a cloud server of different levels, and carrying out cooperative training on heterogeneous models deployed by nodes of different levels in combination with a federated learning method and knowledge distillation; the knowledge distillation is introduced in the federated learning process, so that models with different capacities and architectures can be deployed at different hierarchy nodes, and the compressed sampling value with the internal privacy protection characteristic is used for replacing original data or proxy data as a distillation medium, so that the privacy risk caused by the knowledge distillation is eliminated from the data source. And the communication efficiency, privacy protection and computing power utilization optimization of the end-side cloud are realized.

The invention discloses a Cuckoo search algorithm-based network node interaction method and system, and the method comprises the steps: setting wireless sensor network parameters, constructing a node energy consumption model containing a free space and a multipath attenuation mode, and adaptively selecting an energy consumption mode according to the relation between a transmission distance and a threshold value; selecting a single-hop or multi-hop transmission mode according to the distance between the cluster head nodes and the base station, determining the optimal number of cluster heads, and optimally selecting the cluster head nodes by using a cuckoo search algorithm; and optimizing a data transmission path, and selecting a relay node through a fitness function to realize interaction with a core network control function node. The system comprises a network initialization and parameter configuration module, an energy consumption modeling module, a cluster head selection and optimization module and a path optimization and core network interaction module. According to the invention, network energy consumption can be effectively reduced, network life is prolonged, data transmission efficiency and network adaptability are improved, and intelligent network management is realized.

The utility model belongs to the technical field of AI edge computing, and particularly relates to a high-performance AI box, which comprises a casing, an HDMI (High Definition Multimedia Interface) data input interface, an HDMI data output interface and a power supply interface are arranged on the surface of the casing, and a circuit board is arranged in the casing. The circuit board is provided with a central processor, a data storage module, an HDMI data decoding module, a data transmission module, an FPGA module and a power supply module, and the box is based on an RK3588 processor, is strong in operational capability, supports the efficient operation of a complex AI algorithm, is better in speed and precision of processing algorithms such as face detection, and accords with a high-real-time scene. The communication interface of the high-performance AI box integrates network peripherals such as a gigabit Ethernet, a Wi-Fi module and a 4G module and CAN, RS485 and UART local interfaces, the network adaptability is high, industrial equipment connection is facilitated, and the application range is widened. In the aspect of interaction, the HDMI OUT, the HDMI IN and the audio interface provide multi-element man-machine interaction, video display and audio playing can be realized by accessing the display and audio equipment, and the system is suitable for scenes such as security monitoring and the like.

Methods and apparatuses for network operation adaptation in a spatial domain in a wireless communication system. A method for a user equipment (UE) to report channel state information (CSI) includes receiving: first information related to reception of CSI reference signals (CSI-RSs) on a cell, second information related to determining a first number of CSI reports respectively associated with a first number of operation states on the cell, third information related to triggering a second number CSI reports respectively associated with a second number of operation states from the first number of operation states, and the CSI-RSs based on the first information. An operation state is associated with a configuration for active antenna ports, antenna panels, or transmission reception points. The method further includes determining, based on the second information and the CSI-RSs, the second number of CSI reports and transmitting a channel with the second number of CSI reports.

The application relates to the technical field of power system operation and maintenance training, and discloses a power grid maintenance training system and method based on extended reality, which comprises a cloud rendering and distribution system: the cloud rendering and distribution system is used for rendering an extended reality scene, generating real-time rendering output, and transmitting the real-time rendering output to a terminal device through an edge computing node and a content distribution network; the cloud rendering and distribution system is connected with a real-time dynamic content updating module and a data processing module, is used for receiving real-time state data and student operation behavior data of the device, and is responsible for embedding the data in a rendering scene. Through the cloud-based XR service, the performance requirement of the terminal device is significantly reduced, the cloud rendering and edge computing are combined with the content distribution network, real-time rendering and transmission of high-quality content are realized, in the case of network fluctuation, low delay and smooth experience can be maintained, the network adaptability of the system is improved, and the hardware cost is reduced.

The invention relates to the technical field of data communication, and particularly discloses a data communication method and system of a data communication network, and the method comprises three stages of data preprocessing, coding transmission and receiving recovery. During preprocessing, standardizing the data into a JSON format, and encrypting the data by using an AES algorithm and a dynamic key mechanism; in the coding transmission stage, a coding algorithm is dynamically selected according to network conditions, and a data packet is segmented and transmitted through multiple paths; and the receiving end performs verification, sorting, decryption and format restoration on the data blocks. The dynamic key is generated according to data importance, a transmission environment and a timestamp, and multipath transmission can automatically switch paths. The system is used for realizing the data communication method. According to the invention, the problems of non-uniform data format, insufficient security, poor network adaptability, single transmission path, lack of effective data verification and exception handling mechanisms and the like in the existing data communication method are solved, and the security, reliability and efficiency of data communication are improved.

The invention relates to the technical field of cultural heritage digital protection, and discloses an ancient building traditional culture child propagation display method and system, and the method comprises the steps: carrying out the topological modeling of a child interaction terminal in a target display region, and generating a network adaptation state map; performing semantic migration coding on the network adaptation state map to generate a display resource dynamic fragment index; executing time sequence mapping on the dynamic fragment index through a cross-layer delay tolerance scheduling module, and constructing a resource adaptation instruction list; mapping the resource adaptation instruction table into an edge cache buffer unit, triggering a multi-step redundant breakpoint continuous playing mechanism, and reconstructing a display content distribution path diagram under the weak network; and constructing a display risk response sequence based on a content flash trend critical point appearing in the distribution path diagram, and generating a display interaction entropy leveling instruction. The method has the advantage that the weak network adaptation capability of the child display system in the remote area is improved.

A computer-implemented system and method for distributed memory-native networking using agents. Each agent comprises a unique identifier, payload, memory field containing lineage and policy references, transport metadata, and a cryptographic signature. A modular protocol stack processes these agents through routing, indexing, and consensus layers, with behavior determined by embedded memory. Routing decisions are trust-scoped based on memory-derived access logs and health feedback. Mutation proposals are validated using dynamic, memory-referenced quorum formation. The system enables stateful, policy-bound propagation, structural reorganization, and adaptive execution without centralized coordination or persistent session state. It operates over conventional transport protocols and supports interoperability with cognition-layer payloads, including semantic execution. The architecture allows dynamic trust modeling, autonomous network adaptation, and memory-driven mutation governance across heterogeneous, decentralized, and highlatency environments.

The application provides a low-power wireless self-organizing network dynamic configuration method and system based on adaptive routing, relates to the technical field of network protocols, and comprises the following steps: acquiring node working state parameters, calculating an energy level value to determine a data transmission path, and setting a multi-level data cache occupation threshold; calculating a network load distribution state, and differentially scheduling nodes; selecting a low-power forwarding channel based on node routing applicability; and monitoring a network state in real time and updating parameters. The application improves network adaptability, reduces communication power consumption, prolongs the network life cycle, and improves data transmission efficiency.

This invention discloses a remote monitoring method for the operational status of network-adaptive devices, relating to the field of industrial internet technology. The method includes configuring the data collection points of the monitored devices in a remote monitoring terminal, building a corresponding operation and maintenance rule base for the monitored devices on a monitoring server, and generating a monitoring strategy set. The monitoring strategy set is used to scan the data collection points in real time and sample device status signals. Through jitter reduction and anti-interference processing, stable signals are obtained. These stable signals are then entered into a status determination machine to evaluate signal quality and decision accuracy, generating a device status determination result. This invention is based on a retrieval path reachability status linkage and retransmission strategy, forming a continuous and verifiable status event stream. This balances transmission efficiency with real-time monitoring display, thereby ensuring the integrity, temporal consistency, and reliability of monitoring status in complex networks and high-concurrency scenarios.

The present application relates to the information technology field of knowledge imparting and skill training, and discloses a computer room learning achievement evaluation system based on real-name binding and intelligent collection, which comprises a real-name binding module, a data automatic collection module, a dynamic data matrix modeling module, a dynamic weight optimization module, a comprehensive score output module, a hybrid network adaptation module and a federal learning coordinator; the real-name binding module is used for uniquely binding the learner identity information and the terminal device hardware feature code; and the data automatic collection module is used for collecting the screen operation log, the code compilation result and the training activity space interactive behavior data of the learner end in real time. Through dynamic weight optimization, multi-dimensional data fusion modeling, intelligent network adaptation and visualization technology, the present application solves the defects of subjectivity, interaction delay and strategy vagueness of the traditional system scoring, and realizes the technical improvement of evaluation objectivity, transmission stability and decision accuracy.

The invention provides a sodium ion positive electrode, which comprises two positive electrode active materials with different median particle sizes and two carbon nanotubes with different length-diameter ratios, and meets the following conditions: 1.2 < = D50B / D50A < = 5, 1.5 < = AR2 / AR1 < = 12, 2.1 < = phi < = 9.0, and the structural factor phi meets the formula I. According to the invention, the positive electrode formula design is constrained in a specific phi interval; under different particle size combinations, different CNT length-diameter ratio combinations and different CNT contents, a proper material design window is constructed, so that the synergistic improvement of the slurry and the electrochemical performance is successfully realized.

This invention relates to the field of Industrial Internet of Things (IIoT), and particularly to an edge computing communication method and system for IIoT. The method includes the following steps: Step 1: A user issues a task to a service platform through a user platform. Upon receiving the task request, the service platform classifies and labels the task based on the associated device type and task type, determining the task's real-time level and the importance weight of its associated production links. Step 2: The service platform transmits the classified tasks to the corresponding service sub-platforms. The service sub-platforms obtain historical operating data of the associated devices and basic edge node information, and generate a first computing task carrying task attribute tags based on the classification results, which is then sent to a management platform. Step 3: The management platform manages multiple first computing tasks uniformly. The technical solution of this invention is well-suited to the characteristics of industrial scenarios, taking into account task priority, resource dynamism, and network adaptability.

Embodiments of the present disclosure provide an image processing method and device, a storage medium and an electronic device. The method comprises: obtaining a first image, processing the first image based on a trained image processing model to obtain a second image; wherein the trained image processing model is obtained by training an initial image processing model based on sample data of a first service type, the initial image processing model comprises a trained historical training model of a second service type and at least one to-be-trained network adaptation module inserted in the trained historical training model; the number and insertion position of the at least one to-be-trained network adaptation module in the historical training model are determined based on the first service type. The historical training model is reused to obtain an image processing model adapted to the first service type, which simplifies the training process of the image processing model of the first service type and improves the training efficiency of the image processing model of the first service type.

This invention relates to the field of multimedia communication and network control, specifically to an intelligent network control method and related equipment for low-latency video backhaul. The intelligent network control method includes the following steps: real-time acquisition of key network indicators of the network link, including bandwidth, latency, jitter, and packet loss rate, to provide real-time network environment data support for transmission strategy adjustments. This invention effectively solves the technical challenge of low-latency video backhaul in complex network environments by constructing an intelligent control mechanism that combines network state awareness and video content feature recognition. Specifically, by real-time acquisition of key network link indicators and the introduction of a lightweight CNN model to analyze video content activity, a dual perception capability of network environment and content features is formed, providing data support for dynamically adjusting encoding parameters and achieving a precise balance between video quality and network adaptability.

The application discloses an ultrasonic image offline collection and dynamic synchronous processing method and system, which realizes offline collection and dynamic synchronous processing of ultrasonic images under network interruption conditions through six function modules of network state monitoring, online data synchronization, image collection and processing, offline mode switching, offline diagnosis processing and incremental synchronization. The method comprises the following steps: acquiring network connection state data and determining a working mode; pulling task and template data and caching in an online state; acquiring multi-modal image data and double-writing storage; detecting network interruption and switching to an offline mode; performing AI analysis on the offline collected images and generating a report; detecting network recovery and performing incremental data synchronization. The application solves the business interruption problem caused by the dependence of the traditional PACS system on the network, improves the network adaptability and robustness of the system, reduces the operation and maintenance cost, and shortens the system fault recovery time.

The application provides a cloud application system based on WEB communication interaction technology and a method thereof. The cloud application system based on WEB communication interaction technology comprises a webpage front end, a scheduling server and a cloud virtual machine. The webpage front end is used for receiving user operation instructions, providing interfaces and interactive components for users. The cloud virtual machine is used for running application instances and receiving instructions for execution. The scheduling server is used for information transfer, helping the front end and the cloud virtual machine to establish a handshake connection, and monitoring the running state of the webpage front end. The technical scheme of the application relies on remote desktop technology, has a simple overall structure, does not change the application program logic, and has low development cost. Meanwhile, the system adopts an efficient data transmission format, optimizes network delay, and has good network adaptability. Moreover, the technology itself supports a non-GPU running mode, effectively reducing the cost.

Embodiments of the present application disclose a file fragment uploading method and device, electronic equipment, computer program and storage medium; the embodiments of the present application obtain the file to be uploaded and file information, and the file information at least includes file size;According to the initial network speed of current network, the file to be uploaded is cut, and a plurality of initial file fragments are obtained;Initial file fragments are sequentially uploaded to the server, and the upload time consumption and upload rate corresponding to the initial file fragments are recorded;According to the upload time consumption and upload rate of the uploaded initial file fragments, the size of the next initial file fragment is adjusted, and the target file fragment is obtained;The target file fragment is uploaded to the server. Therefore, by using the mode of "dynamic fragment uploading+network adaptation", in the file uploading process, the real-time network state is determined according to the uploading situation of the uploaded fragment, and the fragment size is dynamically adjusted accordingly, so as to improve the file uploading success rate and user experience, and be flexibly applied to various large file uploading scenes.

The invention discloses an electric energy meter anti-loss method and system based on collection system verification, and the method comprises the steps: achieving the data transmission through the unique hardware coding of a built-in ID chip of an electric energy meter, the encryption through an SM4 cryptographic algorithm, and the communication through an HPLC (High Performance Liquid Chromatography) high-speed power line; a three-level framework of'hardware layer-binding layer-authentication layer 'is constructed, a binding relation between the electric energy meter and the acquisition terminal is established, and a dual authentication mechanism of code matching and position matching is adopted. After the authentication is passed, the electric energy meter activates the meter display and generates a short-term local authorization token, and the meter display can be maintained and the power consumption data can be stored by the token when the network is weak or disconnected; a gradient authentication failure disposal process is matched, if the equipment is lost or embezzled, the equipment returns to a non-screen display state after a token fails due to the fact that a new scene code cannot be matched and an effective activation instruction cannot be obtained, and illegal use value is thoroughly blocked. The problems of passive protection and difficulty in considering protection effect and weak network adaptation in the prior art are solved, and asset safety and power utilization order stability of power supply enterprises are guaranteed.

The present application relates to the technical field of Bluetooth lamps, and more particularly to a Bluetooth lamp remote control system, which first acquires real-time running data of the lamp and dynamically divides control domains; secondly, combines lighting targets and domain data to perform Mesh networking grouping decomposition; then, performs dynamic brightness compensation estimation according to regional brightness nodes and decomposition data; finally, constructs a control dependency graph based on compensation data, quantifies networking conflicts, and generates collaborative control instructions. This method improves network adaptability through dynamic domain division, ensures precise and smooth lighting through brightness compensation, and solves the problems of channel congestion and timing disorder under high concurrency by means of the dependency graph and conflict quantification mechanism, thereby realizing efficient and stable remote dynamic collaborative control of large-scale lighting systems.