Store-and-forward based methods for the signal control problem in large-scale congested urban road networks

Aboudolas, K.; Papageorgiou, M.; Kosmatopoulos, E. · 2009 · Crossref

DOI: 10.1016/j.trc.2008.10.002

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 designing network-wide traffic signal control strategies for large-scale, congested urban road networks. The authors are motivated by the limitations of existing real-time control systems, such as SCOOT and SCATS, which perform well under undersaturated conditions but deteriorate during severe congestion. Furthermore, detailed network models often lead to computationally infeasible optimization problems due to discrete variables. To overcome this, the study utilizes the store-and-forward modeling (SFM) paradigm, which simplifies traffic flow dynamics to allow for efficient optimization while focusing on split optimization to minimize queue spillback. The authors present and compare three methodologies based on SFM: one established method and two novel approaches. The first is a linear-quadratic (LQ) optimal control strategy, which derives a linear multivariable feedback regulator. This method modifies a fixed signal plan in real-time based on queue measurements, using a quadratic knapsack algorithm to satisfy green time constraints. The second novel method is an open-loop constrained quadratic programming control (QPC) approach. It introduces independent link green times as variables to increase control flexibility and explicitly accounts for queue constraints, solving the problem via quadratic programming. The third novel method is an open-loop constrained nonlinear optimal control (NOC) approach. It employs a more accurate nonlinear model that captures intra-cycle traffic dynamics and uses a feasible-direction algorithm to solve the optimization problem, including a penalty term to suppress high-frequency control oscillations. The study evaluates these methods through preliminary simulation-based investigations on a large-scale urban road network. The results demonstrate that all three methodologies are capable of computing signal control plans in real-time. The LQ approach offers simplicity and low computational demand, while the QPC and NOC methods provide greater control flexibility and explicit handling of queue constraints. The simulations confirm the comparative efficiency and real-time feasibility of the developed signal control methods, showing that they can effectively minimize and balance link queues to reduce the risk of oversaturation and gridlock. The significance of this work lies in providing viable, computationally efficient alternatives for managing traffic in saturated urban networks. By leveraging the store-and-forward paradigm, the authors avoid the exponential complexity associated with discrete optimization models. The proposed methods, particularly the novel QPC and NOC approaches, offer improved performance under congested conditions compared to traditional strategies, balancing computational tractability with control effectiveness. This contributes to the field by advancing the design of real-time, network-wide signal control systems that can handle the complexities of large-scale urban congestion.

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