摘要
为了研究城市道路动态变化的交通流量,本文提出了在多分辨率分析(MRA)的基础上利用加乘模糊神经网络(AMFNN)进行预测的方法,并给出了交通流量的预测模型,同时,用实测的流量数据测试了模型的性能。仿真实验表明,采用基于多分辨率分析的加乘模糊神经网络非线性方法能够获得较满意的预测结果。
In this paper, an effective method of data prediction based on wavelet MRA (Multiple Resolution Analysis) and AMFNN (Additive-Multiplicative Fuzzy Neural Network) is proposed to precisely research the changing traffic-flow in urban roads. Accordingly the prediction model is given, and tested by the observed traffic-flow data. Experimental results have shown that the provided nonlinear prediction method has approving performance in the field of traffic-flow.
出处
《交通运输工程与信息学报》
2004年第3期72-76,共5页
Journal of Transportation Engineering and Information
