New traffic models considering complex geometries and data

• Applicants: Professor Dr. Michael Herty; Professor Dr.-Ing. Markus Oeser
• Subject Area: Traffic and Transport Systems, Intelligent and Automated Traffic ; Mathematics
• Term: from 2021 to 2025
• Project identifier: Deutsche Forschungsgemeinschaft (DFG) - Project number 461365406

• Final Report Year: 2025

Final Report Abstract

The traditional approach to traffic safety analysis—based on historical accident data—is limited by the requirement that a sufficient number of accidents must have occurred to identify accident hotspots and implement appropriate measures. The predictive capabilities of the traditional approach are therefore limited and should be supplemented with other methods, such as model-based predictions, or replaced by traffic safety analyses using safety indicators. Traffic modeling provides the opportunity to predict the number of accidents on the studied road section through model-based forecasting. While traffic modeling on highway sections is well advanced, models for more complex geometries (e.g., intersections), which are riskier due to altered driving behavior, are not yet as widely available. As part of the project, the insights and data previously collected on traffic behavior on highway sections were applied to the complex geometries of highway interchanges. The newly developed models and numerical methods were combined with previous results and supplemented with stochastic approaches (uncertainty) to ultimately create a multiscale stochastic traffic forecasting model for the entire highway. Additionally, the safety indicators were adapted and further developed for merging traffic, taking into account the lateral movements and acceleration behavior of vehicles. The project also focused on the further development of a lane-changing model tailored to complex road geometries. The aim was to gain a deeper understanding of the interactions between vehicles on different lanes within traffic flow—particularly in merging scenarios. To this end, safety indicators were systematically investigated, and the model was extended with stochastic parameters to realistically capture varying driving behaviors. The resulting model served as the foundation for its integration into advanced kinetic-stochastic traffic models, applicable to both road segments and junctions. The modeling on the three levels—microscopic, mesoscopic, and macroscopic—was carried out using a calibration process based on microscopic traffic data and driving simulator studies. The microscopic traffic data for road sections and interchanges were partially collected and derived within the project, while some were available from other research projects. The stochastic parameters essential for the modeling were derived based on the analysis of driving simulator studies, data from a weather station, and the collected traffic data.

Publications

• Before-after analysis of temporary traffic regulation in a T-intersection in Aachen, Germany. ICTCT Conference 2022
  E. Kallo, J. Ehlers, A. Pettirsch, S. Lamberty & A. Fazekas
  
• Central schemes for networked scalar conservation laws. Networks and Heterogeneous Media, 18(1), 310-340.
  Herty, Michael; Kolbe, Niklas & Müller, Siegfried
  
• Simulation of surrounding traffic in a driving simulator – Coupling Sumo, RoadRunner and Unity. Sumo User Conference 2022
  R. Schulte Holthausen, M. Berghaus & P.A. Klee
  
• A Central Scheme for Two Coupled Hyperbolic Systems. Communications on Applied Mathematics and Computation, 6(4), 2093-2118.
  Herty, Michael; Kolbe, Niklas & Müller, Siegfried
  
• A Microscopic On-Ramp Model Based on Macroscopic Network Flows. Applied Sciences, 14(19), 9111.
  Kolbe, Niklas; Berghaus, Moritz; Kalló, Eszter; Herty, Michael & Oeser, Markus
  
• Data‐driven models for traffic flow at junctions. Mathematical Methods in the Applied Sciences, 47(11), 8946-8968.
  Herty, Michael & Kolbe, Niklas
  
• Influx ratio preserving coupling conditions for the networked Lighthill–Whitham–Richards model. PAMM, 24(4).
  Kolbe, Niklas
  
• Vehicle trajectory dataset from drone videos including off-ramp and congested traffic – Analysis of data quality, traffic flow, and accident risk. Communications in Transportation Research, 4, 100133.
  Berghaus, Moritz; Lamberty, Serge; Ehlers, Jörg; Kalló, Eszter & Oeser, Markus
  
• Modeling Lane Changes at Freeway On‐Ramps With a Novel Car‐Following Model Based on Desired Time Headways. Journal of Advanced Transportation, 2025(1).
  Berghaus, Moritz & Oeser, Markus