论文部分内容阅读
近年来,短时行程时间预测研究取得了一些成果,研究者在各自的领域内运用不同的方法建立了各种行程预测算法。这些预测算法按照所选参数的多少分为综合模型和基于路段上的时间序列预测模型,综合模型适应性强,但需要的交通参数较多,时间序列预测模型适应性较差,但模型简单。按照预测效果及用途分为实时性预测以及估计行预测,实时性预测要求预测的实时性能高,用来做交通控制、交通流诱导和线路引导。估计性预测其精确度要求不高,实时性能要求也不高,主要用来为出行者提供信息,本文将介绍几种已有的路段行程时间预测算法。
In recent years, some achievements have been made in the study of short-term travel time forecasting. Researchers have used various methods to establish various travel forecasting algorithms in their respective fields. These prediction algorithms are classified into comprehensive models and time-series prediction models based on road segments according to the selected parameters. The comprehensive model is more adaptable but requires more traffic parameters and the time series prediction model is less adaptive but the model is simple. According to the prediction effect and the use, it is divided into real-time prediction and estimated row prediction. The real-time prediction requires high prediction real-time performance and is used to make traffic control, traffic flow guidance and route guidance. Estimation of the accuracy of its forecast less demanding real-time performance requirements are not high, mainly used to provide information for travelers, this article will introduce several existing road segment travel time prediction algorithm.