Agent Based Modelling for Sustainable Public Transport Planning

Abstract

Peripheral regions in Germany are facing significant negative demographic changes, including aging and population decline due to negative natural growth and relocation to industrial centers. Planning urban transport in peripheral regions is a challenging task, as the costs for municipal authorities are high, and despite these costs, users face infrequent and limited departures, long waiting times, and indirect routes. The aim of the Mobidig project, which was financed by the German Ministry of Transport and Digitalization, was to improve transport in these regions. We used demand forecasts and simulations to record and study the mobility needs of the people in the region. Using machine learning and statistical methods we analyzed data on topics such as population distribution, destinations, movements of vehicles and mobile devices, use of means of transport, and basic geodata, and created a virtual image of the region. Based on these analyses, we developed traffic simulation of the city of Hof, which allows us to predict the use of buses under various conditions. This simulation, as a digital twin of traffic, has shown very good alignment with the data used in real traffic. These project results already form the basis for further projects outside and within the region to create an economical and comprehensive public transport system in rural regions.

Mehr zum Titel