基于投影寻踪回归模型的短时交通量预测研究

准确有效地预测短时交通量是实施交通诱导及控制的前提与关键。投影寻踪方法能从不符合正态分布的或没有多少先验信息的数据本身中找出其结构或特征,并能在线性投影中解决非线性结构的问题。文中将投影寻踪回归算法理论应用于短时交通量预测领域,采用正交Hermite多项式拟合岭函数,并用C＋＋语言设计出相应模型的实现算法,对短时交通量实施滚动预测。利用某快速路实际观测数据进行预测实验,实验结果证实该方法具有可行性、可靠性,有一定的实用价值。

The Short-term Traffic Flow Forecast Based on Projection Pursuit Regression

Accurately and efficiently predicting the short-term traffic flow is the premise and key of traffic management and control.Projection pursuit method can find out the structure or characteristics from the data themselves which do not meet the normal distribution or have not prior information,and can figure out the nonlinear structural problems in the linear projection.Therefore,the theory of projection pursuit regression algorithm is applied to forecast short-term traffic flow.The orthogonal Hermite polynomi...