An Empirical Study Of Travel Behavior Using Private Car Trajectory Data

With the development of science and technology,the number of cars in the world is increasing rapidly.Private cars have become the main means of transportation for people.The scope of road construction is constantly expanding,but the traffic problem cannot be solved from the root.However,a massive volume of private car trajectory data generated every day provides a new opportunity for our research.Based on the research of domestic and foreign related literatures on people’s travel behavior,this paper studies people’s travel behavior based on large-scale private car trajectory data.The main work of the paper is as follows:In this paper,we use GPS(Global Positioning System)and OBD(On-board Diagnostic)equipment to collect a large scale of private car trajectory data.The collection equipment and collection process are introduced in detail.In addition,the data are preprocessed to remove invalid,duplicate and redundant data,and the drift data are detected and deleted by threshold method.Moreover,in order to compare the trajectory data of private cars with the trajectory data of taxis and Didi,we introduced the acquisition methods and detailed data of taxi and Didi trajectory data.This paper uses various indicators including the number of trips,average velocity,trip distance and entropy to quantify people’s travel behavior.In addition,we find that private cars have certain regularity in the spatiotemporal pattern through the study of entropy rate.What’s more,we reveal that dwell time is a unique property of private cars,which can identify people’s travel purpose.When using each indicator to analyze the private car trajectory data,we also conducted a comparative study on people’s travel behaviors reflected by the trajectory data of taxi and Didi.The results show that the private car trajectory data is more suitable for mining people’s travel behavior,and the regularities analyzed are more accurate.This paper presents a destination prediction algorithm for private cars,and the algorithm enhances the existing Sub-Trajectory Synthesis(Sub Syn)algorithm.The biggest difference between the two algorithms is that the algorithm proposed by us makes use of the travel regularities of private car to group the trajectories and perform trajectory matching within the group to predict the location of the destination.The experimental results show that the accuracy of the prediction is better than the destination prediction of the existing algorithms,which means that understanding people’s travel behavior can help to predict the destination of private cars more personalized.