小波消噪的神经网络短时交通流量预测模型  

Short Time Traffic Flow Prediction Based on Wavelet De-noising and Neural Network

作　　者：于振洋[1] 

机构地区：[1]淮阴工学院,江苏淮安223000

出　　处：《计算机仿真》2012年第9期360-363,共4页Computer Simulation

摘　　要：研究短交通流量预测问题,短时交通流量数据中含有大量噪声,对预测精度产生不利影响,为了提高短交通流量预测精度,提出一种小波消噪的神经网络短时交通流量预测模型。首先采用小波技术对短时交通流量数据进行消噪处理,然后采用关联维数确定BP神经网络输入变量个数,最后采用BP神经网络建立短时交通流量预测模型。仿真结果表明,与消噪前比,消噪后模型的预测精度有了较显著提高,其预测误差远远小于消噪前,预测结果更具实用价格。The paper put forward a short-term traffic flow forecasting model based on wavelet neural network for. Firstly, the wavelet was usd to carry out the de-noising for the short-time traffic flow data , and then the correlation dimension was used to determine the number of input variables of BP neural network. Finally, the BP neural network model was used as short-term traffic flow forecasting model. The simulation results show that the prediction accuracy of the denoising model is significantly improved, and the prediction error is far less than that of non-denoising.

关 键 词：小波消噪 神经网络 交通流量 相空间重构 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]