基于路网相似性的路段行程时间估计.

Despite the large amount of floating car data,there are still some links lack of real-time data during some certain period of time. Therefore,it is difficult to estimate the travel time. Considering the problem of sparse data when using floating car data estimate the travel time,this paper put forward a kind of inferred method based on big data of floating car. It designed a three-layer artificial neural net-work model,whose input information and output information were the spatio-temporal correla- tion and the target link travel time respectively ,and obtained traffic spatio-temporal correlation relationship using historical big data of floating car and then inferred the travel time of target link. The experiment results indicate that the mean absolute percentage error of travel time can achieve to 30 percent. Compared with Naive model, the accuracy of ANN model performs better, and it is practical to solve the problem of data missing from the aspect of road net-work similarity. [ABSTRACT FROM AUTHOR]