基于GRU神经网络的城市交通流量预测网站设计

Design of Urban Traffic Flow Prediction Website Based on GRU Neural Network

交通流量预测是智能交通系统中必不可少的组成部分。利用门控循环单元(GRU)神经网络搭建交通流量预测模型,并与循环神经网络(RNN)、长短期记忆(LSTM)神经网络进行对比,以验证GRU的优越性。在此基础上,为更好地满足实际应用,实现了基于GRU神经网络的滚动预测,设计了一个城市交通流量预测网站来展示预测结果,方便用户查询,有一定的使用价值。

Traffic flow prediction is an indispensable part of the intelligent transportation system.This paper uses the Gate Recurrent Unit(GRU)neural network to build a traffic flow prediction model,and compares it with Recurrent Neural Net-work(RNN)and Long Short-Term Memory(LSTM)neural network to verify the advantages of GRU.On this basis,in order to better satisfy practical applications,a rolling prediction based on GRU neural network was implemented,and an urban traffic flow prediction website was designed to display the predicted results,which is convenient for users to query and has certain practical value.