The prediction of bus arrival time using Automatic Vehicle Location Systems data

摘要

Advanced Traveler Information System (ATIS) is one component of IntelligentTransportation Systems (ITS), and a major component of ATIS is travel timeinformation. The provision of timely and accurate transit travel time information isimportant because it attracts additional ridership and increases the satisfaction of transitusers. The cost of electronics and components for ITS has been decreased, and ITSdeployment is growing nationwide. Automatic Vehicle Location (AVL) Systems, whichis a part of ITS, have been adopted by many transit agencies. These allow them to tracktheir transit vehicles in real-time. The need for the model or technique to predict transittravel time using AVL data is increasing. While some research on this topic has beenconducted, it has been shown that more research on this topic is required.The objectives of this research were 1) to develop and apply a model to predict busarrival time using AVL data, 2) to identify the prediction interval of bus arrival time andthe probabilty of a bus being on time. In this research, the travel time prediction modelexplicitly included dwell times, schedule adherence by time period, and trafficcongestion which were critical to predict accurate bus arrival times. The test bed was abus route running in the downtown of Houston, Texas. A historical based model,regression models, and artificial neural network (ANN) models were developed topredict bus arrival time. It was found that the artificial neural network models performedconsiderably better than either historical data based models or multi linear regressionmodels. It was hypothesized that the ANN was able to identify the complex non-linearrelationship between travel time and the independent variables and this led to superiorresults.Because variability in travel time (both waiting and on-board) is extremely important fortransit choices, it would also be useful to extend the model to provide not only estimatesof travel time but also prediction intervals. With the ANN models, the predictionintervals of bus arrival time were calculated. Because the ANN models are nonparametric models, conventional techniques for prediction intervals can not be used.Consequently, a newly developed computer-intensive method, the bootstrap techniquewas used to obtain prediction intervals of bus arrival time.On-time performance of a bus is very important to transit operators to provide qualityservice to transit passengers. To measure the on-time performance, the probability of abus being on time is required. In addition to the prediction interval of bus arrival time,the probability that a given bus is on time was calculated. The probability densityfunction of schedule adherence seemed to be the gamma distribution or the normaldistribution. To determine which distribution is the best fit for the schedule adherence, achi-squared goodness-of-fit test was used. In brief, the normal distribution estimates wellthe schedule adherence. With the normal distribution, the probability of a bus being ontime, being ahead schedule, and being behind schedule can be estimated.