Physics-informed deep operator network for traffic state estimation

TL;DR

This paper introduces a physics-informed deep operator network for traffic state estimation that learns to predict full traffic states from limited data while ensuring physical consistency with traffic flow laws.

Contribution

It proposes a novel operator learning framework that incorporates traffic physics directly into the neural network, improving accuracy and physical fidelity over existing methods.

Findings

Outperforms state-of-the-art baselines on NGSIM dataset

Provides insights into optimal function generation strategies

Demonstrates robustness to input data variations

Abstract

Traffic state estimation (TSE) fundamentally involves solving high-dimensional spatiotemporal partial differential equations (PDEs) governing traffic flow dynamics from limited, noisy measurements. While Physics-Informed Neural Networks (PINNs) enforce PDE constraints point-wise, this paper adopts a physics-informed deep operator network (PI-DeepONet) framework that reformulates TSE as an operator learning problem. Our approach trains a parameterized neural operator that maps sparse input data to the full spatiotemporal traffic state field, governed by the traffic flow conservation law. Crucially, unlike PINNs that enforce PDE constraints point-wise, PI-DeepONet integrates traffic flow conservation model and the fundamental diagram directly into the operator learning process, ensuring physical consistency while capturing congestion propagation, spatial correlations, and temporal evolution. Experiments on the NGSIM dataset demonstrate superior performance over state-of-the-art baselines. Further analysis reveals insights into optimal function generation strategies and branch network complexity. Additionally, the impact of input function generation methods and the number of functions on model performance is explored, highlighting the robustness and efficacy of proposed framework.