Skip to content. | Skip to navigation

Sections
Personal tools
You are here: Home Theses Improving passenger demand predictions for public transport with long-term forecasts and real-time occupancy data

Contact

Prof. Dr. S. Decker
RWTH Aachen
Informatik 5
Ahornstr. 55
D-52056 Aachen
Tel +49/241/8021501
Fax +49/241/8022321

How to find us

Annual Reports

Disclaimer

Webmaster

 

 

Improving passenger demand predictions for public transport with long-term forecasts and real-time occupancy data

Thesis type
  • Master
Status Running
Proposal on 07. Sep 2021 13:30
Proposal room Seminarraum I5
Add proposal to calendar vCal
iCal
Supervisor(s)
Advisor(s)

Misallocation of vehicles in public transport networks leads to both overcrowded and unterutilised busses or trains. Both are detrimental to the success of the network as overcrowding is one of the main barriers to adoption and underutilising vehicles is a waste of resources. Resources can be allocated better the more accurate a demand prediction becomes. These predictions can be done either beforehand (without updating the prediction as the day progresses) or in real-time. Real-time approaches that do not depend on manually collected data only consider a small part of the network (e.g. one line) because their approaches lack in scalability.

 

This thesis is aimed at improving upon an established beforehand prediction approach by analysing deviation patterns in real-time. In contrast to prior work, the entire transport network will be taken into account. By delegating the beforehand prediction to an existing approach, it can focus on the deviations. Characteristic patterns in the deviations are extracted and form the basis for predicting deviations in real-time.

Document Actions