关于智能运输系统的关键理论——综合路段行程时间预测的研究

行程时间预测是智能运输系统研究的一个重要问题。为此，建立了许多算法，有历史趋势方法、非参数回归模型、时间序列方法、神经网络、卡尔曼滤波、交通模拟和动态交通分配模型等。然而，在变化的交通状况和任意时段的条件下，这些方法和模型都不能取得令人满意的预测结果。在总结这些已有的预测方法和模型的基础上，提出了一种综合模型。

Study on the Synthetic Link Travel Time Prediction Model of Ke y Theory of ITS

Travel time prediction is one of important problems of ITS research. Some predictive models including historic profile approaches,non parametric regression,time series, neural networks, Kalman filtering, traffic simulation and dynamic traffic assignment models, etc., have been established. None of them can obtain satisfactory predictive results at any time of day and/or under different traffic conditions. On the basis of overviewing the existing prediction models, a synthetic model is proposed.