基于小波神经网络的短时交通流预测

Short-term Traffic Flow Prediction Based on Wavelet Neural Network

针对BP神经网络寻优速度慢、实时性差等缺点,提出将小波神经网络应用于短时交通流量预测中.该方法具有良好的局部性特征,弥补了常规神经网络的不足.在非线性逼近中得到了很好的应用.实验结果表明,小波神经网络预测模型的预测精度明显优于BP神经网络,对交通流量预测问题有较高的应用价值.

Based on BP neural network slow optimization speed and poor real-time faults,Presents the application of wavelet neural network in short-time traffic flow forecasting.The method has a good local characteristics,Make up for lack of conventional neural network.In the nonlinear approximation has a very good application.The experimental results show that Wavelet neural network forecast model is better than BP neural network,It has a high application value on the traffic prediction.