AI Science Tracker
EN

Data-guided equation search finds a flexible empirical car-following model

Tongji University · 20 Oct 2026 · physics

Discovery

Data-guided equation search finds a flexible empirical car-following model

Source check: Source checked. Not independent scientific confirmation.

Finding: Agent-guided and evolutionary symbolic search identified AERM, an empirical equation combining relative-speed response with spacing-dependent restoration. On three driving datasets, its equation fitted complete trajectories more closely than seven classical models when each trajectory was independently calibrated. The result concerns fitted reconstruction, not forecasting unseen drivers or a universal behavioral law.

Technology: Agent-guided symbolic regression; evolutionary equation learning

Publication: Preprint · Test context: retrospective or benchmark. Catalog categories, not an evidence grade.

AI contribution
Data-fitted evolutionary symbolic search and evidence-guided structural proposals generated the equation form; this is not merely parameter fitting of a pre-specified classical model.
Test and result
Equation structure was selected using US-101, highD and exiD; Shanghai, I-80 and Waymo supplied additional reconstruction comparisons with recalibrated coefficients.
Limitation / still unknown
Per-trajectory calibration fits the same complete trajectory subsequently reconstructed. Common-parameter results were mixed and Gipps was better on Waymo. The selected0.1-second delay is not established human reaction time; analytical stability assumes zero delay. The reasoning agent base model is not identified in the inspected primary Methods. No predictive transfer, traffic intervention or causal driver mechanism was established.
Evidence sources and annotation context

Evidence context (descriptions follow the cited sources; test type is a catalog classification):

Evidence sources: https://arxiv.org/abs/2610.27836v1 · https://arxiv.org/html/2610.27836v1

Source checked · editorial details

The source page was checked against this entry. Independent confirmation is not assessed by this label.

Source type: paper. Paper: linked separately.

Review date: 21 Nov 2026. Review scope: Original v1 front, actual structural/model Methods, specific result evaluation and source caveats.

No follow-up recorded; this is not a claim that no follow-up exists. Follow-up review scope and date are not recorded. No source-supported relationships recorded.

Editorial assessment: Impact 2/5 · Validation 2/5 · Novelty 2/5. Impact and novelty are editorial judgments; validation follows the annotated test type where one exists. Not an evidence grade. Scoring criteria.

Open in the full catalog for filters, charts and related discoveries. Browse physics discoveries.