Research On Real-time Detection Of Traffic Information And Bus Arrival Time Prediction Based On Probe Bus

Intelligent Transportation System (ITS) is the trend of development in transportationindustry, and real-time detection in transportation system is very important in ITS.However, the traditional methods to detect have disadvantages in both precision andreal-time detection. So the breakthrough in Probe Vehicle technology comes intoresearchers' view, and have become the hot area recently.Currently, Probe Vehicle research is mostly based on taxi. To overcome thedisadvantages in taxi-based model and sample theory, this paper provides a new wayto sample and process the data based on bus, called Probe Bus Data ProcessingMethod (PBDM). PBDM first shapes a velocity distribution for the samples inverifying period, and then verifies the distribution with the samples in updating period.According to different results of the verification, PBDM treats the sample differently.Results from simulations show that PBDM is effective. Second, this paper discusseson the factors influencing the precision of Probe-Bus detection, including the lengthand volume of the link, and the quantity of Probe-Bus. Then this paper provides therules to judge the practical sample size and the theoretical sample size in a link,defines the concept of bus hot and cold area, and introduces new ways to improve theprecision of detection.Leveraging the PBDM, this paper probes into the prediction of bus arrival time.According to the comparison of point-based, link-based and path-based prediction, theresults from simulation show that link-based and path-based prediction are better. Tomeet the passengers' further need, this paper makes use of link-based and path-basedprediction, and invents an integrated way to help passengers choose the right bus.To simulate the Probe-Bus system, this paper designs and implements the module ofProbe-Bus in Simulation and Analysis System of Urban Mixed Traffic.The arrangement of chapters are as follows:The first chapter lists the disadvantages of traditional methods of detection, andintroduces the Probe Vehicle technology. Then it compares the taxi-based andbus-based methods, and brings in the situation of research on prediction of bus arrivaltime. The second chapter discusses the PBDM, and the influencing factors indetection in details. Also the judging rules of practical sample size and theoreticalsample size, and concept of bus hot and cold area are talked over. The third chapterdiscusses and simulates on point-based, link-based and path-based prediction of bus arrival time, and provides an integrated way to meet passengers' further need. Thefourth chapter focuses on the design and implementation of the Probe Bus module insimulation system. The fifth chapter summarizes this paper and leads to the potentialdevelopment in future.