Coordinated Ramp Metering Control based on Scalable Nonlinear Traffic Dynamics Model Discovery in a Large Network

TL;DR

The paper tackles the challenge of coordinating ramp metering in large highway networks by learning accurate nonlinear traffic dynamics from data and embedding this model into an MPC framework. It combines Koopman with control, SINDYc for sparse nonlinear system identification, and model predictive control to compute coordinated ramp metering actions. Through a SUMO-based case study of a large-scale network with three intersecting highways and eight on-ramps, the proposed SINDYc-MPC framework outperforms ALINEA, PI-ALINEA, and DMD-MPC in reducing occupancy deviations and increasing traffic throughput, while enabling real-time applicability. The work demonstrates a scalable, data-driven pathway to improved traffic management in complex urban networks with potential for broader deployment and extension.

Abstract

This study proposes a coordinated ramp metering control framework in large networks based on scalable nonlinear traffic dynamics model discovery. Existing coordinated ramp metering control methods often require accurate traffic dynamics models in real time, however, for large-scale highway networks, since these models are always nonlinear, they are extremely challenging to obtain. To overcome this limitation, this study utilizes the Sparse Identification of Nonlinear Dynamics with Control (SINDYc) to derive the accurate nonlinear traffic dynamics model from observed data. The discovered dynamics model is then integrated into a Model Predictive Control (MPC) coordinated ramp metering controller, enabling optimized control actions that enhance traffic flow and efficiency. The proposed framework is tested on a large-scale highway network that includes three intersecting highways and eight on-ramps, which outperforms the existing approaches, demonstrating its effectiveness and potential for real-time application. This framework can offer a scalable and robust solution for improving real-time traffic management in complex urban environments.

Figures (7)

• Figure 1: Framework of the proposed data-driven ramp metering model predictive control.
• Figure 2: Structure of the tested highway network including three highways: CA-134E, CA-2S, and I-5N
• Figure 3: Comparison of average traffic occupancy and flow between the no-control case and the local feedback control case.
• Figure 4: Comparison between $\dot{x}$ predicted by SINDYc and $\dot{x}$ through numerical differentiation
• Figure 5: Traffic Occupancy (%) at each sensors of the proposed and benchmark models (moving average is applied to smooth the lines).