Analysis of public transport travel time (on the example of lviv)

Abstract

An important aspect of Intelligent Transport Systems (ITS) is the provision of accurate information on the duration of public transport. By knowing the time of arrival of shuttle vehicles, it is possible to reduce the waiting time for passengers and to attract more people to use public transport. Existing approaches have two major limitations in the field of bus travel forecasting. The first is the large number of factors affecting traffic and the limited amount of real-time travel time data in modern cities make it difficult to accurately predict travel time. The second is mainly the transit time, under the influence of various factors, has different patterns. However, little research has focused on how to divide different routes and build independent models for them. The authors propose a new segmental approach to predicting travel time for public transport, using a model of real bus traffic data that is defined in different segments of bus routes. The authors evaluated the approach using real trajectories collected in Lviv during April 2019. Compared with existing methods, the experimental results show that this approach improves the accuracy of predicting of bus travel time, especially in the case of uneven vehicle flow.