Bus Travel Time Prediction Based On GPS Data

The development of public transport priority strategy is significance to solve the urban traffic problem.In the public transport,the passengers are most concerned about the bus stop time.Bus arrival time forecast will help to enhance the attractiveness of the bus,which is essential to ease the traffic congestion problem.Firstly,the paper introduces the existing forecasting model of bus arrival time and analyzes the characteristics of each model.In the arrival time prediction needs to meet the real-time and practical requirements.The paper is based on historical data to select similar sections to achieve arrival time prediction.The main work of this paper is as follows:1.Using the Newton position-time interpolation method.The GPS data near the site for interpolation,get the site location of the accurate time to rely on the station.The experimental results show that the time taken by the interpolation is accurate and can show the driving characteristics of the bus near the site.2.According to the actual data on the bus operation of the law analysis.To discuss the operation of the bus at different times and days,to determine the main factors affecting the operation of the bus.Through the K-Means clustering method,the running time of historical section is analyzed.We can divide the running state of the section.From the history of the road to select the most similar sections of the calculation.Modeling the realization of the bus in different locations of the data model,on this basis,put forward bus time rolling forecasting model.Based on the similarity of road running status to achieve bus arrival time to predict.Finally,the arrival time of different time periods is predicted and compared with other arrival time prediction models.The experimental results show that it can predict the bus arrival time accurately and meet the travel demand of passengers,and has practical and real-time characteristics.