An MPC-driven VSL control of motorway traffic by employing a continuum second-order traffic model

Variable speed limit (VSL) system driven by model predictive control (MPC) is implemented as a proactive traffic control measure for effective and safer traffic flow operations. Conventionally, the application of the VSL control hinges upon the discrete second-order macroscopic models. Conversely, the MPC paradigm hardly ever employs continuum second-order models for traffic control. In this paper, an improved continuum second-order macroscopic model is used as a predictive model in VSL control driven on MPC. The VSL control is based on actual motorway traffic in Dublin, and is implemented in a microscopic traffic simulation. The findings of the study are two-fold: first, the predictive capability of the continuum second-order model is evaluated; second, the effectiveness of the MPC-driven VSL control strategy is assessed. Within this MPC paradigm, the motorway experience a significant reduction in traffic congestion. The efficacy of the proposed VSL control is demonstrated with a key performance index, flatness index (FI). An average FI of 0.74 is observed, indicating moderate to strong improvement of traffic flow with the VSL control. In other words, 58% to 76% improvement is observed in the congested segments of the motorway. The proposed framework is vital for effective traffic control as it demonstrates the capability to alleviate congestion. • We present an MPC-driven variable speed limit (VSL) control for the Dublin motorway (M50). • The MPC framework utilizes an enhanced continuum second-order traffic flow model. • The accuracy of the traffic flow model is validated by comparing its output with ground truth data generated in SUMO. • The VSL control effectively reduces congestion on the motorway, demonstrating its feasibility for real-world implementation.