Electric power data stream transmission time reasoning method based on graph neural network

Abstract

The invention provides an electric power data stream transmission time reasoning method based on a graph neural network, which is used for solving the problems that the transmission time of each data stream cannot be accurately reasoned and the operation time is long in the existing transmission time reasoning. The method comprises the following steps: establishing a GNN model: establishing a topological structure of the GNN model according to a network structure and routing information of a power data center, and mapping data flow information of the power data center into characteristic values of an attribute graph in the GNN model; carrying out GNN model training: training a GNN model through supervised learning by using the collected data set to obtain a GNN reasoning model; reasoning the transmission time: mapping the test data collected in the power data center into a characteristic value of the GNN model, inputting the characteristic value into the GNN reasoning model, and performig reasoning to obtain the transmission time of the data stream. According to the method, the transmission time of the data stream can be inferred quickly and accurately, decision-making of data stream transmission and scheduling is facilitated, and the operation efficiency of a network in a power data center is improved.

Description

technical field [0001] The invention relates to the technical field of power data centers, in particular to a graph neural network-based power data stream transmission time reasoning method. Background technique [0002] A large number of power services are deployed in the power data center, such as power billing, power government affairs, real-time transactions, etc., which are indispensable and important components for the normal operation of the current power system. How to efficiently use the computing, storage and transmission resources of the power data center is an important issue, which is not only related to the construction income in the power data, but also affects the operating efficiency of the power business. Wherein, the transmission resources of the power data center specifically refer to its network resources, that is, the data center network interconnecting a large number of servers of the power data center, which is also the background technology that the ...

AI Technical Summary

Problems solved by technology

[0007] Aiming at the technical problem that the existing inference method of data flow transmission time cannot accurately infer the transmission time of each data flow and the running time is long, the present invention proposes a graph neural network-based power data flow transmission time inference method, with the help of the latest Artificial intelligence technology --- Graph Neural Network (GNN) to infer the transmission time of data flow in the power data center, and has the obvious advantages of accurate results, fast calculation, and simple use

Images