Bus Arrival Time Prediction Model Study in APTS

Abstract

Setting electronic indicators at bus stations and releasing bus operation information can inform bus passengers of the bus operation situation, so that bus passengers can easily choose their travel habits, bus service can be more efficient, and more passengers will be attracted to use public transportation. In this way, traffic congestion can be eased and air pollution and energy consumption can be reduced. This paper, based on previous studies, studies the contents and methods of the bus arrival time prediction, and establishes the bus arrival time prediction model. This article divides the travel time into three parts: free travel time, road intersection delay time, stop delay time, using the combination of Autoregressive Integrated Moving Average models, the Webster road intersection delay model and the delay model on passengers getting on or off the bus respectively to forecast the travel time. Finally, this paper uses No.649 transit lines of Tianjin as examples to forecast travel time. By comparing the data of bus travel time obtained from the two ways of investigation and models, the effectiveness and accuracy of the established models are examined．