21 results about "Multiple aggregation" patented technology

The embodiment of the invention provides a hierarchical network topology layout method and device based on topology potential, and the method comprises the steps: calculating the topology potential value of a node according to the position information, connection relation and service attribute of a known network node, and positioning a key node based on the topology potential value; performing clustering layered layout on the network nodes according to the key nodes and the service areas where the key nodes are located to obtain a plurality of clustering points; and taking the plurality of clustering points as a first hierarchy of the network topology, visually displaying representative nodes of the first hierarchy, and performing hierarchical topology display on a subnet region where the representative nodes are located based on an interactive operation of a user on the clustering points.

The invention discloses a multi-element aggregation regulation and control method suitable for user-side flexible resources, and belongs to the field of power system operation and control. The multi-aggregation regulation and control method is designed for aggregating the large-scale user side flexibility resource regulation capability to participate in centralized management and control and coordinated scheduling, meanwhile, the system flexibility requirement is considered, and in order to meet the flexible supply and demand balance and weaken the conservative property of inner approximation feasible region calculation under the condition of avoiding an infeasible solution, the method has a good application prospect. According to the method, the flexible resource aggregation feasible region can be accurately described, the optimization effect is not affected while the solving time is remarkably shortened, and the method can be used for solving the flexible resource aggregation feasible region under the constraints of three aspects of the disposable computing power, the storage space and the computing time. And centralized control and coordinated scheduling of large-scale flexible resources are realized.

The invention provides a pipeline overturning lining repair operation process monitoring and evaluation method and device and a computer readable storage medium, through introducing a PELT change point detection algorithm, time sequence data is divided into different time periods based on the detected change point number, and then feature vectors of different time periods are extracted; based on the multi-dimensional feature vectors, the Transform model is adopted to predict the future features of the process parameters, the prediction precision of the Transform model is improved, accurate process parameter monitoring is achieved, a multi-index decision CoCoSo method is introduced, through multi-aggregation strategy fusion, the optimal solution is pursued, the consistency of all strategies is considered, the reliability of the decision result is improved, and the method is suitable for large-scale popularization and application. Therefore, a reliable process state evaluation result is obtained, dynamic monitoring and adjustment of process parameters of overturning lining repair operation are achieved, the repair quality is improved, and safe and stable operation of a gas pipeline system is guaranteed.

A method for generating a spatial model tile data organization structure, relating to the field of geographic information technology, is characterized by the following steps: generating tile network leaf nodes; generating tile network intermediate nodes; selecting an initial aggregation point for the tile leaf nodes and performing multiple aggregation operations with each leaf-level tile until all leaf-level tiles are aggregated to obtain multiple tile aggregates; binding the tile network leaf nodes to their parent tile network intermediate nodes based on the correspondence between the tile network leaf nodes and the aggregated tile network intermediate nodes, ensuring that each tile network intermediate node has a logical relationship with its child nodes; and generating a tile network root node. Advantages: Effectively generates a multi-level detailed tile organization structure for model data in a given data block; the level of detail in each layer of tile data is reasonable, and the data organization is efficient; the data storage scale of a single tile is low and stable, saving network bandwidth and improving file transfer efficiency.

Methods, systems, and devices for wireless communication are described. A user equipment (UE) may determine that a base station has scheduled the UE for one or more downlink transmissions, each of the one or more downlink transmissions having an associated repetition factor corresponding to one of a plurality of configured repetition factors configured at the UE. The UE may identify the applied repetition factor to apply to feedback codebook generation for the one or more downlink transmissions. The UE may generate a feedback codebook for reporting feedback for the one or more downlink transmissions, the feedback codebook being populated at least in part based on the applied repetition factor and based on whether the one or more downlink transmissions are successfully received and decoded. The UE may transmit a feedback report to the base station that includes the feedback codebook.

The present application provides a multi-aggregation node federated knowledge distillation learning method, system, device and medium. The method includes correcting the predicted probability of the client, obtaining the model parameters of the aggregation node and the model parameters of the adjacent aggregation nodes of the aggregation node according to the calculated corrected probability; using the data volume of the public data set of each aggregation node as the weight, calculating the weighted average of the model parameters of the aggregation node and the model parameters of the adjacent aggregation nodes of the aggregation node to obtain the updated model parameters of the aggregation node and the aggregation node update model; according to the aggregation node update model, obtaining the predicted probability of the aggregation node update model on the public data set and correcting it; based on knowledge distillation, training the personalized model of the client connected to the aggregation node according to the calculated corrected probability. Public data sets are established separately by multiple aggregation nodes to share the communication pressure and storage pressure.

A technique for managing metadata changes in a storage system provides different aggregation policies for use with different metadata. The technique includes assigning metadata changes to respective aggregation policies and storing the assigned metadata changes in a set of data structures. The technique further includes destaging the metadata changes from the set of data structures separately for the different aggregation policies in accordance with settings specific to those aggregation policies.

The application discloses a kind of object group determination method, device, equipment, storage medium and program product, involve artificial intelligence and data processing technical field.The method comprises: obtaining the aggregation zone structure graph corresponding to reference object group, and the multiple aggregation zone structure graphs corresponding to the network to be detected;Obtain the distributed representation corresponding to reference object group, and the multiple distributed representations corresponding to the network to be detected, the distributed representation is used to represent the structural features of aggregation zone structure graph;According to the distributed representation corresponding to reference object group, determine the target aggregation zone structure graph from the multiple aggregation zone structure graphs corresponding to the network to be detected;Obtain the target object group corresponding to target aggregation zone structure graph.The application determines the object group by supporting the aggregation zone level, without determining object one by one, and considering the overall correlation of object group, so as to improve the determination efficiency and determination accuracy of object group.

The present invention discloses a real-time indicator system for bank flow computing business based on multi-dimensional intermediate state aggregation, in which a preprocessing module receives real-time transaction flow from a Kafka message queue and interacts with a full database Redis module. The feature processing module screens and combines the features of the preprocessed data to generate intermediate state records. The indicator aggregation module is composed of a plurality of aggregation computing nodes, and the intermediate state records are dispersed to different computing nodes for aggregation according to the aggregation method. The aggregation operation process uses a parallel computing method and can use the underlying computing resources for acceleration, and the final calculation results are stored in the intermediate state database MemDB module. The present invention can improve the efficiency of computing resource utilization, balance the computing load, and effectively reduce the system IO load, so as to meet the real-time requirements of bank flow computing business indicators.

A computer system displays a graphical user interface that includes a data visualization region, a shelf region, and a schema information region. The computer system receives a first user input to place a first dimension data field from the schema information region into the shelf region. In accordance with the first user input, the computer system determines a visualization level of detail that is based on at least the first dimension data field. The computer system receives a second user input to specify a first aggregation. The first aggregation has a custom level of detail that is distinct from the visualization level of detail. The computer system builds a visual specification that includes the visualization and custom levels of detail. The computer system generates a data visualization corresponding to the visual specification and displays the data visualization in the data visualization region.

The invention relates to a file pre-distribution method and device, computer equipment and a storage medium. The method comprises the following steps: acquiring file access information in a preset time period and a plurality of preset aggregation algorithms corresponding to each resource file, and performing aggregation processing on access parameters of the resource files by adopting the plurality of preset aggregation algorithms corresponding to the resource files to obtain aggregation data corresponding to different preset aggregation algorithms, and fusing the aggregation data corresponding to different preset aggregation algorithms, determining a file pre-distribution list based on a fusion result, and issuing the resource files to a target node according to a to-be-distributed number in the file pre-distribution list, thereby completing pre-distribution of the resource files to edge nodes. Based on the method, multiple aggregation algorithms are combined in the file pre-distribution process to allocate the resource files for the edge nodes, a more comprehensive and more accurate pre-distribution result can be obtained, and therefore the file pre-distribution requirements in different scenes can be met.

The application provides a kind of virtual power plant task scheduling processing method, electronic equipment and storage medium, belong to power distribution network control technical field.The method comprises: receiving to be scheduled task, to be scheduled task includes target power and power deviation threshold;According to to be scheduled task, select multiple aggregation units with different aggregation types and distribution positions from the regulation unit library as the first unit, and issue the first scheduling instruction of executing target power to the first unit;Select multiple aggregation units matched with the aggregation type and distribution position of the first unit from the margin unit library as the second unit, and issue the second scheduling instruction of executing power deviation threshold to the second unit;According to the response result of first scheduling instruction and second scheduling instruction, the first unit that does not satisfy the first condition is adjusted to the margin unit library, and the second unit that satisfies the second condition is adjusted to the regulation unit library.The application improves the test dynamicity and accuracy of aggregation unit response deviation degree.

To assist with replanting work. [Solution] The afforestation system includes a processor. The processor sets up multiple aggregation areas in an afforestation area where multiple plants are planted, calculates a numerical value related to plant death for each aggregation area, and generates display data showing the relationship between the aggregation areas and the numerical values ​​of the aggregation areas.

The present application relates to the technical field of motion pattern recognition, and provides a motion pattern mining method based on mobile target trajectory and related equipment. The method includes: obtaining the distance change sequence of each mobile target pair based on all trajectory data; calculating the convergence rate and separation rate according to all distance change sequences, and determining multiple convergent mobile targets and multiple separated mobile targets from all mobile targets according to the convergence rate and separation rate; clustering the convergent mobile targets according to the trajectory data at the current moment to obtain multiple convergent clustering clusters, and clustering the separated mobile targets according to the trajectory data at the first moment to obtain multiple separated clustering clusters; taking the movement of all mobile targets in each convergent clustering cluster within a time period as a convergent pattern, and taking the movement of all mobile targets in each separated clustering cluster within a time period as a separated pattern. The method of the present application can improve the versatility and reliability of motion pattern mining.

The present invention relates to a printed circuit board defect detection method and device based on cross-scale multiple aggregation, and relates to the field of target detection. The method comprises: obtaining a printed circuit board defect image and performing preprocessing; using an improved YOLOv8 network to perform feature extraction and fusion, wherein the feature extraction network adopts CDA-net, and the network comprises an embedding layer module, a four-layer CDA module and an SPPF module, and can effectively extract multi-scale features; the feature fusion network adopts RGS-FPN to perform efficient multi-scale feature fusion; an anchor-free YOLOv8 detection head is used for target detection, and an Adam optimizer is used to train the model with a learning rate of 0.001; the preprocessed image is input into the trained model to obtain the defect detection result of the printed circuit board. The present invention improves the accuracy and efficiency of printed circuit board defect detection through deep learning technology, and is suitable for quality control in the manufacturing process of printed circuit boards.

The present application is applicable to the technical field of network resource management, and provides a data transmission method and related equipment for an air-space-ground integrated network, including clustering sensing nodes to obtain multiple aggregation nodes; constructing a first data transmission model between aggregation nodes and sensing nodes; constructing a second data transmission model and a third data transmission model between multiple aggregation nodes and multiple drones, and between multiple drones and low-orbit satellites; constructing a first optimization model based on the first and second data transmission models; constructing a second optimization model for reducing drone data backlog based on the third data transmission model; constructing a third optimization model based on the first and second optimization models, and obtaining a data transmission strategy for configuring an air-space-ground integrated network through alternating iterations of the first and second optimization models; and realizing data transmission of the air-space-ground integrated network based on the data transmission strategy. The present application can ensure the timeliness of end-to-end data transmission in the air-space-ground integrated network.

The invention discloses a traffic matrix calculation method and device, a storage medium and electronic equipment, and the method is applied to the field of network traffic detection, and comprises the steps: collecting the node information of each node in a target network, and determining the network information of a plurality of aggregation domains according to the network information of the target network and the node information of each node; expanding the target information according to the network information of the plurality of aggregation domains, and constructing a target mapping relationship between each node in the target network and the plurality of aggregation domains according to the expanded target information; collecting traffic data of each node in the target network through a distributed traffic collection tool; and calculating an aggregation traffic matrix of the target network according to the target mapping relation and the traffic data of each node. Through application of the method and the device, the problem that the response speed of a service system is relatively low due to the fact that the calculation load and the storage demand of the traffic matrix are increased in a square level along with the expansion of the network scale in a service scene of real-time traffic matrix updating in related technologies is solved.

The application discloses a kind of multistage cache aggregation computing method and device, it is related to cache aggregation computing technical field, the method includes: connecting message bus, obtains the first level cache corresponding to stack top;When subscribing to the data of same time dimension and device dimension, cumulative aggregation calculation is carried out with original data;Whether the first event task belongs to the time interval defined in range is judged;When the timing task corresponding to the first event task triggers, set delay waiting time;At the same time, the data of the first level cache is transmitted to external cache middleware, and the data of same time dimension, device dimension is updated and merged.The application solves the technical problems of single cache mode, limited aggregation mode, low data storage and processing efficiency in the prior art, improves the efficiency of time series data processing, reduces memory usage and delay, supports multiple aggregation modes, and enhances the technical effects of system performance and scalability.

The invention provides a virtual power plant task scheduling processing method, electronic equipment and a storage medium, and belongs to the technical field of power distribution network regulation and control. The method comprises the following steps: receiving a to-be-scheduled task, wherein the to-be-scheduled task comprises target power and a power deviation threshold value; selecting a plurality of aggregation units with different aggregation types and distribution positions from an adjustment unit library according to the to-be-scheduled task as first units, and issuing a first scheduling instruction for executing target power to the first units; selecting a plurality of aggregation units matched with the aggregation type and the distribution position of the first unit from the margin unit library as second units, and issuing a second scheduling instruction for executing a power deviation threshold to the second units; according to response results of the first scheduling instruction and the second scheduling instruction, the first units which do not meet the first condition are adjusted to the margin unit library, and the second units which meet the second condition are adjusted to the adjustment unit library. According to the invention, the test dynamics and accuracy of the response deviation degree of the aggregation unit are improved.

The invention discloses a hierarchical cooperative control method for multiple aggregation distributed resource clusters of source-load storage. The method comprises the following steps: dividing control levels of the multiple aggregation distributed resources of the source-load storage; based on the control level division result, aggregating the resources of the same control level into the same resource cluster, and constructing a plurality of aggregated distributed resource clusters of source-load storage; establishing a distributed resource cluster cooperative control model; and introducing a graph neural network model to solve the distributed resource cluster cooperative control model. According to the method, an improved fuzzy clustering algorithm is designed to classify resources, a mechanism of weighted information propagation and dynamic adjustment of propagation strength is introduced to improve an information transmission mechanism of the graph neural network, and the graph neural network is innovatively applied to the solution of a distributed resource cluster cooperative control model; and the modeling capability and the solving efficiency of a complex system are improved. And the calculation efficiency is kept in a high-dimensional and nonlinear scene, and the operation requirements of a large-scale distributed power grid are met.

A federated learning method applied to the field of artificial intelligence technology. In the method, when the aggregation node obtains models of different network structures, the models of the same network structure are divided into the same group, and the parameters of the models in the same group are aggregated to obtain multiple aggregation models with different network structures; and for each aggregation model, knowledge distillation training is performed on each aggregation model based on the multiple models obtained originally, to realize experience transmission between models of different network structures, thereby integrating the knowledge and experience of models of various network structures, combining the advantages of parameter aggregation and knowledge distillation in integrating model experience, realizing aggregation of models of different network structures, and ensuring the prediction accuracy of the model obtained after aggregation.