Network traffic flow optimization under performance constraints

Csikós, Alfréd; Charalambous, Themistoklis; Farhadi, Hamed; Kulcsár, Balázs; Wymeersch, Henk · 2017 · Crossref

DOI: 10.1016/j.trc.2017.08.002

archive: archived pipeline: cataloged verified

Get this paper ↗ (DOI — opens at the source; we link to it, we don't host it)

Summary

This paper addresses the challenge of optimizing network traffic flow in large-scale urban vehicular networks under specific performance constraints. Motivated by the negative impacts of urban congestion, such as delays and emissions, the authors propose a model-based perimeter control policy that manages vehicle admission into a protected network. The approach utilizes the Network Fundamental Diagram (NFD) to model aggregated network dynamics. Unlike traditional methods that focus solely on maximizing throughput, this work incorporates two critical performance metrics: maintaining average network travel time below a defined threshold and preventing the blockage of external queues (vehicles waiting to enter). The primary contribution is a control strategy that jointly handles these requirements, transforming an originally non-convex optimization problem into a numerically efficient convex one by relaxing performance constraints into time-dependent state boundaries. The methodology models the system using vehicle conservation equations for both internal (protected network) and external (entrance queues) states. The network outflow is defined as a concave function of network accumulation via the NFD. The control objective is to maximize network throughput while satisfying hard constraints on average time delay and external queue capacity. The authors derive upper and lower bounds for the internal queue length based on these performance requirements. Specifically, the travel time constraint provides an upper bound on internal accumulation, while the external queue capacity constraint provides a lower bound. This reformulation allows the optimization problem to be solved as a single-step receding horizon control with convex constraints. When arrival rates are too high to satisfy both constraints simultaneously, the method prioritizes maintaining the travel time threshold, accepting external queue blockage to preserve internal network performance. The proposed method was evaluated through comparative numerical simulations using the microscopic traffic simulator Vissim, modeling a portion of Stockholm’s inner city. The study compared the proposed "relaxed controller" against a Proportional-Integral (PI) controller, a Nonlinear Model Predictive Control (MPC) approach, and an uncontrolled fixed-time strategy. Results demonstrated that the PI controller, while effective at minimizing travel time, failed to prevent external queue blockages. The MPC controller struggled when performance constraints conflicted, violating both delay and blockage thresholds. In contrast, the relaxed controller successfully prioritized constraints, keeping travel delays within bounds even when external queues reached capacity. The simulation confirmed that the method effectively manages the trade-off between internal network efficiency and external queue management, detecting and handling scenarios where simultaneous satisfaction of all constraints is impossible. The significance of this work lies in its ability to provide a real-time, computationally efficient solution for perimeter control that explicitly accounts for Quality of Service metrics like travel time and queue stability. By converting complex non-convex problems into convex ones with time-varying bounds, the approach offers a robust mechanism for urban traffic management. It ensures that prescribed performance requirements are met under varying demand conditions, offering a superior alternative to existing PI and MPC strategies that either ignore constraints or lack clear prioritization logic during conflict scenarios. This contributes to the field by integrating communication network service indicators into traffic control, enhancing the reliability and predictability of urban transportation systems.

Provenance

The full processing record for this entry. Every stage of this paper's journey through the pipeline is logged — what ran, with which tool and model, how many attempts it took, and when it last completed.

Summary generated by qwen3.6-27b-prismaquant on 2026-06-26; verification: verified.

Topics

Ranked by relevance to this paper. Hover a topic for its definition.

• traffic density