VLR (Visitor Location Register) user number predicting method based on gray system model neural network

Abstract

The invention discloses a VLR (Visitor Location Register) user number predicting method based on a gray system model neural network, comprising the steps of obtaining a VLR user number in an area in a plurality of months, and five indexes influencing the VLR user number in telecommunication business, wherein the indexes comprises mobile company market share, a mobile telephone popularizing rate, festival and holiday days in each month, disposable income of residents, and number of resident population; building the gray system model neural network; and training and testing the gray system model neural network, and predicting the VLR user number in a town by the gray system model neural network passing tests. The method is capable of accurately and rapidly predicting telecommunication business volumes under the conditions that known information is little, and part system characteristic knowledge is lacked, and can well reflect the fluctuation of the telecommunication business volumes due to population mobility and other factors.

Description

technical field [0001] The invention discloses a novel telecommunication service prediction method, in particular a VLR user number prediction method based on a gray system model neural network. Background technique [0002] Telecom business forecasting is a precondition for the phased construction planning of the communication network, and it is also one of the necessary conditions for telecommunication business volume and investment estimation during the planning period. The choice of forecasting method is directly related to the accuracy of forecasting results. There are many traditional forecasting methods, such as trend extrapolation method, growth curve method, etc., but the forecasting results of these forecasting methods have the following problems: the forecasting results are all smooth curves, and it is impossible to intuitively describe the telecommunication services caused by seasons and other external environments. fluctuations: [0003] 1. The prediction resu...

AI Technical Summary

Problems solved by technology

There are many traditional forecasting methods, such as trend extrapolation method, growth curve method, etc., but the forecasting results of these forecasting methods have the following problems: the forecasting results are all smooth curves, and it is impossible to intuitively describe the telecommunication services caused by seasons and other external environments. fluctuations:

[0003] 1. The prediction results are all smooth curves, which cannot intuitively describe the volatility of telecom services

[0004] 2. The forecast results do not respond significantly to changes in important factors such as national policies and market strategies

[0005] 3. Other neural network algorithms tend to fall into local optimal solutions

Images