19results about How to "Relieve computing pressure" patented technology

The invention belongs to the technical field of information, and discloses a digital twinning-based NOMA-assisted cloud side end communication computing method, which comprises the following steps of: constructing a cloud side end cooperative communication computing architecture, considering the problems of terminal isomerism and edge server load balance, partially unloading a computing task, and selectively transmitting the unloaded part to a local edge server for computing. Or selecting a local edge server as a relay node, further transmitting to a third-party edge server or a cloud server for cooperative computing, considering the security problem of edge computing, formulating the problem, constructing an optimization target of resource allocation and task unloading decision and a corresponding constraint condition, and converting into a multi-agent MDP; and the MADDPG algorithm is used for solving in the digital layer, real-time interaction with the physical layer is realized to complete strategy issuing, cloud side-end collaborative communication calculation is realized, the security and accuracy of the problem model are improved, and the Qos demand of the terminal is met.

The invention discloses a new energy consumption scheduling method considering an electric vehicle. The method comprises the following steps: establishing a day-ahead optimization model; performing optimization solution on the day-ahead scheduling by adopting a distributed algorithm to obtain a day-ahead scheduling scheme; constructing an intra-day optimization model; and according to the load and the photovoltaic prediction error value, an intra-day optimization scheduling model strategy is implemented, and rapid response is carried out by adjusting regional power grid interaction, an electric vehicle and energy storage so as to eliminate the error to the maximum extent. By means of the energy storage system and full play of the flexibility of the electric vehicle cluster, the power between different connecting lines can be flexibly adjusted, and it is ensured that new energy can be stably consumed. Meanwhile, load prediction errors need to be considered, and in order to enhance the reliability and stability of the system, the operation of closely tracking the scheduling plan is executed. The rolling optimization strategy can effectively cope with the uncertainty of the day-ahead stage, ensure the smooth consumption of new energy, and reduce the influence of prediction errors on the operation of the system.

The invention discloses a cloud edge hierarchical collaborative information system integrated resource scheduling system. The system comprises an acquisition module. Data preliminary cleaning and feature extraction are locally executed through the edge end, only key feature data are uploaded, network transmission quantity is reduced, cloud computing pressure is relieved, real-time tasks can respond quickly, task timeout caused by transmission delay is avoided, feature vectors of the edge end and the cloud end are dynamically analyzed based on a scheduling decision model of reinforcement learning, and the real-time performance of the edge end is improved. Task requirements and resource states are intelligently matched, the scheduling accuracy is improved, model parameters are dynamically updated through task execution state data (such as duration and resource utilization rate deviation), the model is adaptively optimized along with environment changes (such as edge end hardware aging and cloud load fluctuation), the efficient scheduling capacity is kept for a long time, and the scheduling efficiency is improved. Layered processing and intelligent scheduling avoid edge end overload or cloud end resource idling, resource allocation of a CPU, a memory and the like is optimized, resource waste is reduced, and the system collaboration efficiency is integrally improved.

The application discloses a detection method for a QUIC protocol denial of service attack (QDA) in a backbone network. The method can realize detection of a composite denial of service attack under the QUIC protocol. The detection method comprises an offline training stage and an online detection stage. In the offline training stage, according to the attack characteristics of the composite QDA, representative one-way traffic feature data is extracted, and corresponding feature groups are constructed. These features can effectively deal with the case that there is a large amount of one-way traffic in the backbone network node. Finally, a specific machine learning method is used for training to obtain an attack detection model. The application can be used for QDA attack behavior detection and early warning in a massive traffic scene such as a backbone network, and provides a basis for maintaining the security of the network.

The application relates to a cloud control-based centralized vehicle queue predictive cruise control method and device. The method comprises the following steps: based on a received predictive cruise control request, acquiring real-time position and leading vehicle state information of a vehicle queue, determining an optimal reference speed curve of the vehicle queue according to the real-time position, the leading vehicle state information, a preset fuel consumption model and a preset dynamics model, taking the curve as an input of a centralized queue speed planning algorithm to obtain an optimal driving speed of the vehicle queue through optimization solving, and sending the optimal driving speed to the vehicle queue to realize vehicle queue cruise control. Thus, by combining the advantages of the cloud, the problems of limited prediction range and insufficient information acquisition capability of the existing vehicle queue predictive cruise are solved, the safety, economy and efficiency of the queue predictive cruise control are realized, the problem of gradually increasing energy consumption of the following vehicle caused by grouping is improved, the computing pressure at the vehicle end is relieved, and the energy-saving and system safety and stability boundary capability of the queue is improved.

The invention discloses a game assisting method, an intelligent glasses system, equipment, a medium and a product, the intelligent glasses system comprises intelligent glasses, the intelligent glasses comprise an intelligent glasses body, and a camera module, a main control chip and a multi-mode prompting unit are carried on the intelligent glasses body; the camera module is used for collecting video data of a game scene; the microphone array is used for capturing audio data of a game scene; the main control chip is used for generating game auxiliary information based on the video data and the audio data; and the multi-mode prompting unit is used for performing multi-mode prompting on the game player according to the game auxiliary information. According to the embodiment of the invention, the multi-modal data is collected through the intelligent glasses, game assistance is carried out according to the multi-modal data, auxiliary software is not needed, and the adaptation cost is reduced. Meanwhile, the intelligent glasses display the game picture and the auxiliary tool interface at the same time, so that the game player does not need to frequently switch the sight line between the game picture and the auxiliary tool interface, and the game concentration is improved.

The invention discloses a table question and answer method and device based on a large language model, a medium and a program product, and the method comprises the steps: firstly, receiving a table file uploaded by a user through a file uploading interface, vectorizing the content of the table file, and storing the vectorized content into a vector library; after a question request sent by a user is received, determining a target column corresponding to the question request in the table file by using the large language model; determining a target row corresponding to the question request according to vector similarity between character strings obtained by splicing the question request and the index column of each row of the table; constructing a question sub-table corresponding to the question request by using the target row and the target column; and finally, based on a large language model, according to the question request and the question sub-table, outputting a reply result. According to the method and the device, the large language model can accurately answer the large-volume table content under the condition that the maximum processing length is limited.

The invention relates to a network teaching video knowledge point classification extraction method and device, equipment and a medium, and relates to the field of video processing, the method comprises the following steps: using a preset character recognition model to extract text data of each video key frame in a video key frame sequence to generate a text sequence corresponding to each video key frame; inputting a text sequence corresponding to the video key frame into a knowledge point classification model trained to a convergence state, and generating a corresponding knowledge point label for each video key frame; and based on the timestamps corresponding to the video key frames and the knowledge point tags, cutting the network teaching video into a plurality of video clips, and merging the video clips corresponding to the same knowledge point tag to determine the network teaching video corresponding to each knowledge point tag. According to the invention, the network teaching video corresponding to the required learning and teaching content can be quickly and accurately found, and the learning and teaching efficiency is improved.

The invention discloses a monitoring system and method for network security protection equipment of a central dispatching master station, and relates to the field of power dispatching control, and the method comprises the steps that a security protection equipment data acquisition module inputs collected data into a security protection intelligent control module; a fault judgment module in the security intelligent control module performs fault judgment based on the text data, and sends fault information to an alarm indication module if a fault exists; a fault pre-judgment module in the security intelligent control module performs fault pre-judgment based on the historical log file, and if a fault exists, a predicted alarm date is calibrated and sent to an alarm indication module; and a network attack detection module in the security intelligent control module performs attack prediction based on the network flow information, and sends abnormal information to an alarm indication module if the abnormal information exists. According to the system, operation monitoring personnel are helped to organize and carry out emergency disposal work leisurely, the influence of equipment faults on a service system is greatly reduced, safe and stable operation of a large power grid is further guaranteed, and the influence of the equipment faults on the service system is effectively reduced from the source.

The invention relates to the technical field of artificial intelligence, in particular to an AI function extension method and system based on the deep learning technology, and the method comprises the steps: obtaining node information of a plurality of network nodes in the same network, extracting corresponding node function features from the node information, and constructing a node function table; performing clustering analysis on the functions of all the artificial intelligence terminals in the network nodes to obtain node functions of the network nodes; constructing a function vector matrix matched with the node function table, and obtaining AI model data matched with the node function features through the function vector matrix; a preset network node connection strategy is imported, network node connection is initialized, the network node connection strategy is adjusted and updated through a preset deep learning model, and the running state of each network node is predicted; and expanding the AI function of one or more network nodes according to the connection condition of the network node.

This invention relates to the field of transportation monitoring technology and discloses a method for synchronous acquisition of transient waveforms in a distributed monitoring terminal for transformer transportation impacts. The key technical points are: deployment of monitoring terminals, wireless time synchronization, mechanical wave compensation, transient waveform acquisition and uploading, and impact analysis; adopting an IoT synchronization method combining LoRa wireless communication time synchronization and mechanical wave compensation, the method involves the slave device to acquire impact waveforms in segments and send impact messages exceeding the threshold in real time, and the slave device completes signal acquisition, filtering, and data processing before uploading to the host, thereby realizing synchronous acquisition of transient waveforms in a distributed monitoring terminal for transformer transportation impacts.

The invention provides a fault power distribution network power supply recovery determination method and device and electronic equipment, and the method comprises the steps: constructing a fault power distribution network power supply recovery optimization model according to a plurality of first constraint conditions; decomposing the optimization model according to a decomposition algorithm to obtain a fault power distribution network main optimization model and a plurality of normal power distribution network sub optimization models; solving the main optimization model and the plurality of sub optimization models to respectively obtain a first objective function value, a second objective function value and an objective function value, and determining the maximum value of the objective function value of the optimization model according to the sum of the objective function values of the sub optimization models and the first objective function value; determining the minimum value of the objective function value of the optimization model according to the sum of the objective function value of the sub-optimization model and the second objective function value; and comparing the difference between the maximum value and the minimum value with a preset threshold value, and recovering the power supply of the fault power distribution network according to conditions. According to the method, the power supply recovery rate is effectively improved on the premise of controlling the cost.

The present invention discloses a method, system and device for edge computing resource sharing based on coalition game, which includes that an edge server collects all device nodes and their task information in the network to form an edge computing network; setting a coalition structure and establishing an initial single-device coalition structure C i , where i ∈ 1, 2, …, n; modeling the utility function of the device nodes; setting the coalition game and the preference operation relationship, and selecting an improved coalition structure C j , and calculating the coalition revenue V(C j ); obtaining the stable coalition structure set C * , until no coalition structure changes, stop the iteration; reaching the stable coalition structure set C * , obtaining the overall system offloading strategy, and the game reaches equilibrium. It can maximize the task completion rate of all device nodes, solve the problem that all tasks are submitted to the edge server resulting in task blocking and processing failure, relieve the computing pressure caused by its resource limitation, and at the same time enable all device nodes to obtain the maximum benefit, improving the overall efficiency and benefit of the system as well as the satisfaction of all parties.

The invention discloses a spatial hybrid modeling method for snow water equivalent raster data based on landform characteristics. The method comprises the following steps: S1, collecting data and preprocessing the data; s2, dividing all the preprocessed data according to landform division; s3, screening out influence factors for different areas; s4, aiming at each divided area, taking the snow water equivalent as a dependent variable, taking the screened factors influencing the snow water equivalent change as independent variables, and respectively constructing a space Durbin model; s5, judging whether the spatial Durbin model residual error of each region has spatial autocorrelation or not, and if yes, turning to step S6; otherwise, taking the parameters of the spatial Durbin model as the parameters of the final model of the region; s6, constructing a spatial error model for the residual error of the spatial Durbin model; s7, merging a space error model result into a space Durbin model to obtain a final space hybrid model; and S8, model precision evaluation and analysis. The method improves the precision of the model.

The present invention discloses a business report generation method, an electronic device and a storage medium, which are used to solve the technical problem that the existing method of generating business reports through calculation views occupies a large amount of memory and easily causes system freeze or dump. The present invention comprises: receiving a deadline for a preset target business input by a user; obtaining a calculation view framework corresponding to the target business; obtaining business data based on the deadline and the calculation view framework, and extracting at least one business data table from the business data; generating a business report of the business data based on the business data table.

The application discloses a multi-layer multi-scale grid and a dynamic auxiliary positioning method based on the same, and belongs to the field of wireless positioning. The method comprises the following steps: performing grid division on a target GNSS reference station network coverage area to obtain longitude difference dl, latitude difference db and height difference dh of the target GNSS reference station network coverage area grid; determining a multi-layer multi-scale grid of the target GNSS reference station network coverage area; based on the multi-layer multi-scale grid, a data center receives rough position information sent by a user, searches for a grid point closest to the user as a positioning service grid, and sends information of a virtual observation station of the positioning service grid to the user, so that real-time positioning is completed. The application solves the centimeter-level positioning demand of the service end for a large number of users from the principle of VRS technology, overcomes the problems of low terminal RTK fixing rate and poor accuracy in a complex terrain area, reduces the calculation pressure of the service end, has simple algorithm and strong practicability, and effectively reduces the construction and operation and maintenance cost caused by simply relying on the accumulation of server and other hardware facilities.

The invention provides a power quality disturbance monitoring method and device, electronic equipment and a storage medium, and belongs to the technical field of power quality monitoring. The invention discloses a power quality disturbance monitoring method. The method comprises the following steps: constructing a mathematical model of a standard signal and a disturbance signal related to power quality; constructing a disturbance data set by using the electric energy quality mathematical model; constructing a power quality disturbance identification model; training a disturbance recognition model by using the disturbance data set; power grid operation data are collected in real time, and the disturbance recognition model is used for monitoring the electric energy quality. By monitoring the power quality disturbance, the disturbance event can be found in time, the characteristic parameters such as the disturbance amplitude and the duration time can be determined, and a basis is provided for improving the power quality.

The invention discloses a double-rate model-free control method driven by a high-frequency motor. The core principle of the double-rate model-free control method is that a calculation-intensive task is separated from high-rate control execution; specifically, complex online model identification and time delay compensation are executed through slow-rate interruption, so that an accurate data driving model without physical parameters is obtained; meanwhile, a control sequence of a high-rate control period is generated at a time in slow-rate interruption by using an interpolation time translation prediction method; then, the high-rate interruption is only responsible for executing the pre-calculated control sequences one by one in sequence; due to the fact that correction based on the latest identification model is achieved on each high-speed control point, and the calculation burden of high-speed interruption is extremely low, the method fundamentally solves the problems of calculation real-time constraint and model parameter dependence in high-frequency control, and therefore high-robustness quasi cycle-by-cycle control of the high-frequency motor driving system is achieved.

This invention relates to the field of spatial distribution management technology for power distribution circuit elements, and discloses a digital map system integrating BIM and GIS dynamic element synchronization functions. The system extracts the pose increment data of distribution circuit node entities in the local space of the building information model, and calculates the scale projection coefficient based on their physical reference coordinates in the geodetic coordinate system. The pose increment data is weighted and mapped by the scale projection coefficient to generate a standardized pose vector, and the spatial landing point is corrected by combining electrical topology constraint rules. This invention effectively solves the spatial tearing and cumulative drift problems caused by cross-scale mapping, realizes the accurate alignment between the logical connection state and physical spatial pose of distribution node entities, and ensures the physical authenticity of the topology structure in power system monitoring.