Trajectory Investigation for Enhanced Calibration of Microsimulation Models
archive: archived pipeline: cataloged verified
Get this paper ↗ (full text — opens at the source; we link to it, we don't host it)
Summary
This report addresses the challenge of calibrating microsimulation models, which are critical for analyzing transportation alternatives and predicting congestion but often suffer from inaccurate representations of microscopic driver behavior. Traditional calibration relies on macroscopic inputs like aggregated speed and throughput, which can produce models that fit aggregate metrics but fail to replicate realistic vehicle dynamics such as car-following and lane-changing. Motivated by advancements in unmanned aerial vehicle (UAV) technology and data processing, the study aims to develop and validate a novel methodology for calibrating these models using vehicle trajectory data. The research team collected large-scale trajectory datasets at four congested urban freeway sites: I-270 in Maryland, I-15 in California, I-95 in Virginia, and I-75 in Florida. Data were gathered in spring 2019 using drones at three sites and a helicopter at I-75. The team processed video footage into numeric trajectory data, implementing rigorous cleaning and validation procedures to correct errors. They developed a seven-step, software-agnostic trajectory-based calibration methodology involving inputs, heuristics, outputs, points, binning, pairing, and root-mean-square error (RMSE) calculations. This method was compared against traditional calibration and a hybrid approach using both macroscopic and trajectory data. Experiments were conducted using two microsimulation software platforms to evaluate the accuracy of the resulting models. The results demonstrated that traditional calibration methods are unreliable for producing realistic vehicle trajectories, even when macroscopic performance measures appear accurate. Explicit integration of trajectory data into the calibration process significantly improved the realism of simulated driver behaviors. The trajectory-based calibration performed best at the I-75 site, where helicopter collection enabled the acquisition of full-length trajectories approximately 1.2 miles long. In contrast, drone-collected data at other sites were limited to 800-foot segments due to height restrictions, suggesting that shorter trajectory snippets may be insufficient for robust calibration. Validation tests confirmed that models calibrated with the new methodology passed accuracy checks after necessary adjustments, whereas traditional methods failed to capture microscopic dynamics. The study concludes that trajectory-based calibration offers a superior alternative to traditional methods for ensuring microsimulation models accurately reflect local driver behavior. The report provides detailed instructions, software scripts, and lessons learned for data collection, processing, and calibration, aiming to facilitate adoption by state and local transportation agencies. The authors recommend future improvements in data collection technologies to capture longer trajectories cost-effectively and urge software developers to integrate user-friendly tools for trajectory-based calibration. This work establishes a foundation for more reliable traffic simulation, with potential applications in studying connected and automated vehicles and calibrating models for different congestion regimes.
Key finding
Explicit integration of vehicle trajectory data into the calibration process remedies the inability of traditional macroscopic calibration methods to produce realistic vehicle trajectories.
Methodology
field_study
Provenance
The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed. Discovered via bulk_ingest_rosap on 2026-05-23 (6 acquisition events logged).
| Stage | Outcome | Tool | Model | Prompt | Attempts | Completed |
|---|---|---|---|---|---|---|
| discover | success | rosap | — | — | 2 | 2026-05-23 |
| archive | success | — | — | — | 1 | 2026-05-23 |
| extract | success | cached | — | — | 2 | 2026-06-10 |
| clean | success | — | — | — | 1 | 2026-06-01 |
| chunk | success | — | — | — | 1 | 2026-06-01 |
| embed | success | — | — | — | 1 | 2026-06-02 |
| enrich | success | — | — | — | 1 | 2026-05-23 |
| promote | success | — | — | — | 1 | 2026-05-23 |
| summarize | success | llm | qwen3.6-27b-prismaquant | summ-v5 | 3 | 2026-06-10 |
| tag | success | vector_similarity | — | — | 24 | 2026-06-11 |
| verify | success | — | — | — | 2 | 2026-06-10 |
Summary generated by qwen3.6-27b-prismaquant on 2026-06-10; verification: verified.
Topics
Ranked by relevance to this paper. Hover a topic for its definition.
Information type
What kind of knowledge this paper contributes, grouped by family — independent of topic (what it is about) and method (how it was studied).
- Methodological Resource: validation psychometrics, tool software
- Theoretical Contribution: computational model