动态交通流量预测方法研究

随着智能运输系统的广泛应用,实时交通流量预测的重要性也日益显著。本文介绍了预测模型发展过程中比较重要的几个模型,并由此引出人工神经网络。介绍误差逆传播(BP)模型的相关理论。指出传统BP神经网络的缺陷,并提出提高预测精度的措施引进高阶神经网络。建立普通BP神经网络的预测模型,利用误差反传播算法实现这些影响因素到输出变量的复杂映射,再用高阶神经网络构建另一预测模型。利用交叉口实测数据进行预测,并用实际数据进行比较验证。

Research on Method of Dynamic Traffic Flow Forecast

With the widely use of the Intelligent Transportation System, the position of the Real-Time Dynamic Traffic Flow Forecast is becoming more and more important. This paper introduces some important models in the development of the forecasting model and brings in the Artificial Neural Network (ANN). Then this paper introduces the correlative theory of the back propagation (BP) model and points out the shortcomings of traditional BP network and some improvements are put forward ring in the High-order Generalized Neural Network. First, construct the neural network forecasting model, with the algorithm of error back propagation, model can reach the complex mapping from these factors to variable output. And then, the other forecasting model is built by High-order Generalized Neural Network (HGNN). Last, with the data of the traffic flow in the intersection, the both kinds of neural network models in real-time traffic flow forecast are run. Then compare the results with the real data.