Based On The Time Distribution Of Bus Travel Time Prediction Model Research

The present transit scheduling level of our country is still in the experience scheduling mode. We can not control the real-time situation of public transport timely. Through analyzing the reason for the laggard engineering level in transit scheduling, It comes the conclusion that the bottleneck of our transit scheduling lies in how can we easily, economically and efficiently access to bus link travel time. This paper has a systematic and in-depth research for bus link travel time. It can not only improve the theory of transit scheduling, but also signify transport dispatchers.First of all, several typical travel time prediction methods will be introduced in this paper. Then, based on buses in Jinan Shandong province and probe car travel time acquisition method based on GPS, the whole travel time prediction model is established through analyzing the data of travel time. At the same time, exploring the factors that may influence travel time and the travel time parameter model will be established. At last, according to the test for this model, the thesis can find that the travel time prediction model has a higher forecast accuracy and practicality.In the process of establishing the travel time prediction model, for the bus operating characteristics, different prediction model is used in different period of time. Meanwhile, travel time parameter model is established with the factors influencing travel time. This model changes the way of transit scheduling in the past, and also has a better practical value.