Multi-Modal Traffic Simulation Calibration and Integration with Real-Time Hardware in Loop Simulator

on YouTube ↗20m

" summary

Built a GPS-based calibration framework for Woody Hayes Drive using OSU CABS data, dividing the network into 12 sections and optimizing nine entry-point generation probabilities with a genetic algorithm stopped at 10 generations under a PMRE objective. Implemented a SUMO+dSpace ASM co-sim with a 5 Hz SUMO / 100 Hz dSpace split, sending ego pose, returning an ROI of 10 vehicles and extrapolating across ~200 ms SUMO gaps. Demo on the OSU campus runs end-to-end in real time.

AI-written summary · about

" chapters (23)

1. 0:00Introduction
2. 0:27Background
3. 1:05Objective
4. 1:48Agenda
5. 2:07Data Source
6. 3:18Traffic Count
7. 3:40GPS Data
8. 3:58Travel Time Calculation
9. 5:05Turn Ratios
10. 6:05Pedestrians
11. 6:40Calibration Setup
12. 8:34Traffic Algorithms
13. 9:15Optimization Function
14. 10:20Genetic Algorithm
15. 11:15Errors
16. 11:29Assumptions
17. 12:22Loop Simulation Framework
18. 13:40Sumo Network
19. 14:31Communication Interface
20. 15:21Simulation Time
21. 15:49extrapolation method
22. 16:38cosimulation
23. 17:34Conclusion