摘要
针对传统的应用数学模型方法在短时交通流预测精度和实时性方面存在的问题,提出了将Volterra滤波器自适应预测模型用于短时交通流的实时预测。为提高预测精度,在Volterra滤波系数计算过程中采用归一化最小均方自适应算法进行多次训练。应用该预测模型对几个典型的非线性系统进行预测,验证了算法的准确性。然后再用此方法对微观实测交通流的时间序列进行实证分析。仿真结果表明,该预测模型对实测交通流时间序列具有很好的预测效果,可以满足实时交通流预测的需要。
Directed at the problems of the prediction precision and real time problem using mathematical model for short- time traffic flow, a real-time adaptive forecasting model for short-term traffic flow based on nonlinear Volterra filter is presented. The optimal Volterra filter coefficients are updated by using normalized least mean square （NLMS） adaptive algorithm in order to improve prediction. Firstly, the time series of several typical nonlinear systems are predicted by the method in order to confirm the veracity of results of the algorithm. Secondly, time series of real traffic flow are researched with it. The simulation results show that the proposed method has effective prediction results for real traffic flow, and it can completely meet the need of real-time prediction of traffic flow.
出处
《系统工程》
CSCD
北大核心
2009年第9期60-64,共5页
Systems Engineering
基金
国家自然科学基金资助项目(50478088)
河北省科学技术研究与发展项目(07276933)
