Enhancing traffic flow predictions with environmental models

In recent years, the Intelligent Transport System (ITS) has been developed, which aims at traffic flow forecasting. It also plays an important role in the planning and development of traffic management and control systems. By analyzing up-to-date traffic flow data, a prediction of the traffic flow is to be made. A precise forecast model can help to manage traffic proactively and to reduce congestion as well as possible traffic accidents.

This traffic flow prediction project is proposed with environmental models for traffic networks. Nowadays traffic flow prediction mainly takes into account information from individual, specific sensors. However, information from neighboring sensors and other sensors in the traffic subnet could be used to improve modern forecasting models.

In this proposal, two dimensions of the traffic flow are involved: first, the geo-factor which contains additional data from neighboring traffic intersections and second, traffic flow behavior. In the evaluation phase of the project, the accuracy of the forecast for a regional transport network is increased.

“I used the SDIL platform for my master’s thesis. Previously, I attended an SCC seminar to learn more about the features of the platform and how to use it. In my opinion, it is very helpful for students who would like to write their thesis in cooperation with the SDIL,” says Qianqian Cao, former student of the Karlsruhe Institute of Technology. “During my master’s thesis, the SDIL team helped me quickly and easily with problems. For example, HTcondor was still new to me at the beginning.”

Data Innovation Community

Smart Cities

Project partners

Siemens, KIT

Contact persons

Daniel Weitze, daniel.weitze@siemens.com

Andreas Hapfelmeier, andreas.hapfelmeier@siemens.com

Project duration

Feb. 2017 – Jul. 2017