Bus Travel Time Prediction Based On Combined Model Of Kalman Filtering And Exponential Smoothing

With the rapid growth of economy and society,the number of private cars purchased by urban residents has also increased greatly.At the same time,traffic congestion and shortage of resources are also brought.Many problems of urban road traffic in China have become more and more important factors affecting the rapid development of modern cities.Time prediction can help improve the operation efficiency of urban public transportation system and reduce the cost of bus operation.It is also an important development direction of the intelligent bus system.In order to improve the service quality of the urban road bus system and alleviate the pressure of urban road traffic resources,many researchers have carried out a lot of investigation and Research on the travel time prediction of the bus.This paper first introduces the research background and current situation of bus travel time prediction,and then introduces some commonly used forecasting models.In view of some of the problems,this paper mainly has done the following work:(1)in this paper,the travel time of the bus is divided into site stopping time and inter Station Road time.Many studies before the time of inter Station Road travel are generally based on the driving information of the target line bus,and the unexpected events and bus self are not considered well in the time data acquisition.In this paper,the data preprocessing method based on the bus travel time of the target line is proposed,which is based on the bus travel time of the target line.It is very good to deal with the problem that the single line bus is not enough to reflect the real time traffic flow,and improves the accuracy of the prediction results.At the same time,In this paper,an improved K-means clustering algorithm is proposed to cluster the historical berthing data of the bus according to the site.(2)the basic contents of the two prediction models of Kalman filtering algorithm and exponential smoothing method proposed in this paper are described in detail.According to the characteristics of the two algorithms,the combination model of the Kalman filtering algorithm and the exponential smoothing method is proposedto predict the travel time between the buses,and the advantages and groups of the combination of the two algorithms are combined.The combination principle and combination method are explained in detail.(3)This paper selects the Taishan Road Station of the Changjiang street from the 236 line bus route to the North Hospital Station as the experimental section,and makes a clustering analysis on the stopping time of the middle Shahezi community and the ninth printing plant,and combines the information of 190 and 106 roads in the prediction of the travel time between the stations and through the experiment.The accuracy and real-time of the combination of multi line bus information fusion and Kalman filter and exponential smoothing method are better.