Quantifying Uncertainty In Traffic State Estimation Using Generative Adversarial Networks

TL;DR

This paper introduces PhysGAN-TSE, a GAN-based physics-informed deep learning framework that quantifies uncertainty in traffic state estimation by integrating stochastic traffic flow models, demonstrating improved robustness on real-world data.

Contribution

It develops a novel PhysGAN-TSE framework combining GANs with physics models for uncertainty quantification in traffic state estimation.

Findings

PhysGAN-TSE outperforms pure GAN and traffic flow models in uncertainty quantification.

ARZ-based PhysGAN achieves better performance than LWR-based model.

The method shows robustness on the NGSIM dataset.

Abstract

This paper aims to quantify uncertainty in traffic state estimation (TSE) using the generative adversarial network based physics-informed deep learning (PIDL). The uncertainty of the focus arises from fundamental diagrams, in other words, the mapping from traffic density to velocity. To quantify uncertainty for the TSE problem is to characterize the robustness of predicted traffic states. Since its inception, generative adversarial networks (GAN) have become a popular probabilistic machine learning framework. In this paper, we will inform the GAN based predictions using stochastic traffic flow models and develop a GAN based PIDL framework for TSE, named ``PhysGAN-TSE". By conducting experiments on a real-world dataset, the Next Generation SIMulation (NGSIM) dataset, this method is shown to be more robust for uncertainty quantification than the pure GAN model or pure traffic flow models. Two physics models, the Lighthill-Whitham-Richards (LWR) and the Aw-Rascle-Zhang (ARZ) models, are compared as the physics components for the PhysGAN, and results show that the ARZ-based PhysGAN achieves a better performance than the LWR-based one.