Potential failures of physics-informed machine learning in traffic flow modeling: theoretical and experimental analysis

TL;DR

This paper analyzes why physics-informed machine learning can fail in traffic flow modeling, highlighting issues with data resolution, gradient alignment, and residual sampling, and provides theoretical bounds explaining observed performance differences.

Contribution

It offers a theoretical and experimental analysis of PIML failures in traffic modeling, identifying key conditions affecting model accuracy and residual behavior.

Findings

Low-resolution data impairs gradient-based updates.

Physics residuals lose effectiveness due to sampling and averaging.

Higher-order models have larger residual bounds, affecting performance.

Abstract

This study investigates why physics-informed machine learning (PIML) can fail in macroscopic traffic flow modeling. We define failure as cases where a PIML model underperforms both purely data-driven and purely physics-based baselines by a given threshold. Unlike in other fields, physics residuals themselves do not hinder optimization in this setting. Instead, effective updates require both data and physics gradients to form acute angles with the true gradient, a condition difficult to satisfy with low-resolution loop data. In such cases, neural networks cannot accurately approximate density and speed, and the constructed physics residuals, already degraded by discrete sampling and temporal averaging, lose their ability to capture PDE dynamics, which directly leads to PIML failure. Theoretically, although LWR and ARZ solutions are weak solutions, for piecewise $C^k$ initial data they remain $C^k$ off the shock set under mild conditions, which has Lebesgue measure zero. Thus, almost all detector or collocation points lie in smooth regions where residuals are valid, and the MLP's inability to exactly represent discontinuities is immaterial. Finally, we establish MSE lower bounds of physics residuals: higher-order models such as ARZ have strictly larger consistency error bounds than LWR under mild conditions. This explains why LWR-based PIML can outperform ARZ-based PIML even with high-resolution data, with the gap shrinking as resolution increases, consistent with prior empirical findings.