交通流量的混沌特性分析及预测研究

Chaotic Analysis and Prediction for Traffic Flow

对城市交通流进行了混沌识别,结果表明其具备混沌特性,结合神经网络和混沌理论,提出一种将混沌理论和神经网络相结合的交通流预测模型,利用重构相空间的嵌入维数确定神经网络的结构,算法与设计相对简单。实例证明了该方法的正确性和科学性,具有较强的应用推广价值。

The paper finds chaos character in traffic flow and puts forward the new forecast model for traffic flow ,which＇ based on neural network and chaos theory. By taking advantage of structures of neural network being determined by embedding dimension of phase space reconstruct, this method is proved to be easy .Practical data also show that the method is effective, scientifieal and involves great practical importance.