Communication bandwidth forecasting method and device based on power business requirements

Abstract

The invention discloses a communication bandwidth forecasting method based on power business requirements. The communication bandwidth forecasting method includes: calculating the total business bandwidth data of each year; obtaining the data of a plurality of business bandwidth influencing factors in each year; performing dimension reduction process on the data of the plurality of business bandwidth influencing factors in each year by means of a principal component analysis method; training a neural network model used in forecasting of an electric power communication network; and forecasting the communication bandwidth of the electric power communication network through the trained neural network model. Additionally, the invention further discloses a communication bandwidth forecasting device based on power business requirements. The neural network model is used for forecasting multiple factors of the bandwidth of the electric power communication network. The problem that the influence on forecasting objects caused by the various business bandwidth influencing factors cannot be described by the traditional method is solved, and the forecasting accuracy is improved. The communication bandwidth forecasting method is favorable for reducing communication cost and accordingly improves the economic benefit and social benefit of the electric power system communication.

Description

technical field [0001] The present invention relates to the technical field of power communication network bandwidth prediction, in particular to a communication bandwidth prediction method based on power service requirements and a communication bandwidth prediction device based on power service requirements. Background technique [0002] With the development of smart grid, the development of grid intelligence will realize the transformation of large-scale processes such as grid production, operation, management, and operation to intelligence. More information collection and monitoring points will be deployed in each link of the grid. Volume and bandwidth requirements will continue to rise. Bandwidth allocation, flow control, etc. in communication networks all depend on network service bandwidth prediction to varying degrees. [0003] For power ICT business bandwidth forecasting, the selection of forecasting methods is directly related to the realization of forecasting goal...

AI Technical Summary

Problems solved by technology

When dealing with the prediction problem considering the influencing factors, there are both deterministic trends and certain random influences. It is necessary to separate the deterministic trends, the calculation is more complicated, and it is necessary to assume the zero mean and stationarity of the separation residuals

The current forecasting method generally adopts the time series analysis method, but the time series analysis method cannot describe the influence of various factors on the forecast object, so there is a problem of low forecasting accuracy

Images