Neural Operators for Boundary Stabilization of Stop-and-go Traffic

Yihuai Zhang,Ruiguo Zhong,Huan Yu

2023-12-16

Abstract:This paper introduces a novel approach to PDE boundary control design using neural operators to alleviate stop-and-go instabilities in congested traffic flow. Our framework leverages neural operators to design control strategies for traffic flow systems. The traffic dynamics are described by the Aw-Rascle-Zhang (ARZ) model, which comprises a set of second-order coupled hyperbolic partial differential equations (PDEs). Backstepping method is widely used for boundary control of such PDE systems. The PDE model-based control design can be time-consuming and require intensive depth of expertise since it involves constructing and solving backstepping control kernels. To overcome these challenges, we present two distinct neural operator (NO) learning schemes aimed at stabilizing the traffic PDE system. The first scheme embeds NO-approximated gain kernels within a predefined backstepping controller, while the second one directly learns a boundary control law. The Lyapunov analysis is conducted to evaluate the stability of the NO-approximated gain kernels and control law. It is proved that the NO-based closed-loop system is practical stable under certain approximation accuracy conditions in NO-learning. To validate the efficacy of the proposed approach, simulations are conducted to compare the performance of the two neural operator controllers with a PDE backstepping controller and a Proportional Integral (PI) controller. While the NO-approximated methods exhibit higher errors compared to the backstepping controller, they consistently outperform the PI controller, demonstrating faster computation speeds across all scenarios. This result suggests that neural operators can significantly expedite and simplify the process of obtaining boundary controllers in traffic PDE systems.

Machine Learning,Systems and Control

What problem does this paper attempt to address?

This paper aims to solve the "stop - and - go" instability problem in traffic flow. Specifically, the author proposes to use neural operators to design boundary control strategies to alleviate these instabilities in congested traffic flow. The Aw - Rascle - Zhang (ARZ) model, which is a model containing two coupled second - order hyperbolic partial differential equations (PDEs), is adopted in the paper to describe traffic dynamics. Although traditional boundary control methods such as the backstepping method are effective, their design processes are complex and time - consuming, and they need to solve the complex backstepping control kernel problem. To overcome these challenges, this paper proposes two different neural operator learning schemes to stabilize the traffic PDE system:

1. **The first scheme**: Embed the gain kernel approximated by the neural operator in the predefined backstepping controller.
2. **The second scheme**: Directly learn the boundary control law.

Through Lyapunov analysis, it is proved that under certain approximation accuracy conditions, the closed - loop system based on the neural operator is practically stable. To verify the effectiveness of the proposed method, the author carried out simulations and compared the performance of the two neural operator controllers with that of the PDE - based backstepping controller and the proportional - integral (PI) controller. The results show that although the error of the neural operator approximation method is higher than that of the backstepping controller, they show faster calculation speeds in all scenarios and are always superior to the PI controller. This indicates that the neural operator can significantly accelerate and simplify the process of obtaining the boundary controller for the traffic PDE system.