基于小波神经网络的短时交通流预测

针对现阶段城市道路交通短时交通流预测精度不高的局限性,将小波变换引入到城市道路交通预测过程中,提出一种基于小波神经网络的预测方法。运用美国加州高速公路通行能力度量系统数据作为数据来源,应用小波变换和BP神经网络相结合对其进行预测,然后对预测结果数据进行分析,并对短时交通流进行综合评价。实验表明,该方法与传统的BP神经网络相比较,在短时交通流预测方面具有较好的有效性和优越性。

Short-time Traffic Flow Prediction based on Wavelet Neural Network

Aimed at the limitation of low prediction accuracy at the present stage of city road traffic, in the time of introduced the wavelet-transform into the urban road traffic prediction, proposes a prediction method based on wavelet neural network. Use Caltrans Performance Measurement System data as data sources, combining wavelet-transform and error Back Propagation neural network as a method of predicting, then analysis and evaluation the prediction results, a large number of experiments show that, comparing with the traditional Back Propagation neural network, this method has more effectiveness and superiority in short-term traffic flow prediction.