Automated Calibration of Traffic Demand and Traffic Lights in SUMO Using Real-World Observations
summary
Automated pipeline converts historical counts and signal logs from Ingolstadt into a calibrated 10×10 km SUMO network (≈12,000 edges, 5,600 junctions, 120 signalized intersections; data for 75 intersections). It parses a provider reference topology, matches inbound roads/lanes and loop detectors, emulates hourly fixed‑time signals from measured green ratios, and runs a two‑step assignment (OD matrix → route pool → iterative route pruning to match counts). Implemented in Python with SUMO and geopandas, it fully matches 57/75 intersections, yields median signal‑ratio errors 1.6–1.9% and traffic‑count median errors 13–18%, and requires some manual detector fixes and future adaptive lights.
AI-written summary · about
chapters (8)
- 0:00Introduction
- 0:15Safe
- 0:56Calibration
- 1:37Matching Applications
- 2:48Transfer of Signal Groups
- 6:31Traffic Assignment
- 10:05Results
- 13:34Conclusion
related talks
19mSensor-based Flow Optimization on connected real-world intersections via a SUMO Feature Gap Analysis
14mAnalysis and Modelling of Road Traffic Using SUMO to Optimize the Arrival Time of Emergency Vehicles
24mTowards Smarter Intersections Through Digital Twins - The Coupling of SUMO and SCENIMINI
16mTopology-Preserving Simplification of OpenStreetMap Network Data for Large-scale Simulation in SUMO
13mUtilizing Vehicle Trajectory Data to Characterize Performance Measures of a Highway Corridors
19mSimulation of DACH-Style Detector-Actuated Traffic Signal Control in SUMO
21mDevelopment, calibration, and validation of a large-scale traffic simulation model: Belgium network
19mCycle-Level Distribution-Based Calibration of Microscopic Intersection Models Using UAV Trajectories