基于神经网络的城市轨道交通短期客流预测研究

The Research of Short Term Passenger Flow Prediction Based on Neural Network in City Rail Transit

该文结合神经网络来研究城市轨道交通中短期客流预测问题。设计出了基于自回归神经网络的轨道交通客流预测模型、模型描述及其模型训练算法。通过matlab仿真实验来验证预测模型的性能，优于将最小二乘支持向量机与离散一维Daub4小波分析结合起来预测效果。

This article is to study the short term passenger flow prediction based on neural network in city rail transit.And the pas-senger flow prediction model based on recurrent nerual network in city rail transit,mathematical descripition and training algo-rithm of the model is designed.and the author testifies the performance of prediction mode according to Matlab simulation exper-iment,result show that the passenger flow prediction method based on recurrent nerual network has higher accuracy than wave-let analysis based SVM in short term passenger flow prediction for city rail transit.