GRNN神经网络在铁路客运量的预测研究

Prediction of Railway Passenger Volume Based on GRNN Neural Network

当前国内铁路运输行业发展迅速,需要对客运量进行准确的预测以达到合理地调度。本文以青岛市为例,根据2005年～2017年的资料,建立GRNN广义回归神经网络预测模型,将预测模型计算结果与实际对比,分析相对误差率。研究表明:GRNN神经网络预测模型误差较低,具有良好的预测效果,有利于铁路客运量的预测分析。

At present, the domestic railway transport industry is developing rapidly, so it is necessary to forecast the passenger volume accurately in order to achieve reasonable dispatch. Taking Qingdao City as an example, according to the data of population and GDP of the whole city from year 2005 to 2017, in this paper the GRNN generalized regression neural network prediction model is established, compares results of the prediction model with the actual situation are calculated, and the relative error rate is analyzed. The research shows that the prediction model of GRNN neural network has low error and good prediction effect, which is beneficial to the prediction and analysis of railway passenger volume.