Using Data From The Automatic Vehicle System And The Automatic Fare Collection System To Analyze Transit Service Reliability

Giving priority to the development of urban public transit is the main way to solve the traffic congestion, urban pollution, and meet the increasing demand of motorization. Transit service reliability is an important factor in determining the quality of bus services and the mode choices of passengers. Studying the transit service issues is of much significance for capturing the characteristics of bus operations, enhancing the attraction of urban transit, and improving the efficiency of operations. A large number of historical data from the intelligent transit systems characterized by automatic vehicle location system (AVL) and automatic fare collection system (AFC) provide the new data environment for transit service reliability research and bring the opportunity for exploring the reliability issues in multiple levels and multiple aspects.This paper aims to use data from intelligent transit systems to investigate three contents: evaluating the transit reliability, analyzing the factors affecting transit reliability,and reliability control. The research framework is based on the theory of transit service reliability and the data analysis methods were used in this study.The significant issues of present research in transit reliability were firstly proposed on the basis of a comprehensive literature review and a wide survey of practices. Guided by the research requirements, the pre-processing flow of AVL data was established according to the features of historical data from intelligent transit systems of Chengdu.Secondly, with the goal of comprehensively evaluating transit reliability, this paper analyzed the travel time reliability and headway evenness at the stop, route, and network levels. Besides, the calculation methods of the indictors and the relationship of them in the data environment of intelligent systems were discussed based on a numerical example.The contributing factors affecting travel time variability were modeled using multiple regression models. The results show that factors including the number of traffic signals, bus stops, types of land-uses, provision of bus lane, time periods, passenger activities, and passenger activity variations were found to affect the travel time variability in different models. Specially, the results suggest that the provision of bus lane is more efficient for reducing travel time variability in peak hours than other periods.A mathematical model for the optimal stopping design of limited-stop bus service was developed to solve the problem of unreliability service.The object function is to minimize the total cost of passenger and bus operators' cost. The vehicle capacity and delays caused by bus queuing at stops are considered. Also, the link travel time is assumed to be stochastic to better reflect the reality,This paper contributes to making a depth understanding on the rules and the effect mechanism of transit reliability,and provides the references for decision makings in urban transit operation.