A robust algorithm to solve the signal setting problem considering different traffic assignment approaches

Adacher, Ludovica; Gemma, Andrea · 2017 · Crossref

DOI: 10.1515/amcs-2017-0057

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 Global Optimization of Signal Settings and Traffic Assignment (GOSSTA), a problem that integrates traffic signal control with route choice behavior. The authors argue that traditional approaches often treat signal settings and traffic assignment independently, failing to account for the feedback loop where signal timing influences driver route choices, which in turn affects network congestion. The research aims to minimize total network delay by optimizing signal parameters (cycle time, green split, and offset) while simultaneously considering traffic assignment. The study validates a robust optimization algorithm, the Surrogate Method (SM), against established techniques like Particle Swarm Optimization (PSO) and Genetic Algorithms (GA). The proposed solution employs an iterative procedure that alternates between updating signal settings based on fixed traffic flows and updating traffic assignment based on fixed signal settings. The core optimization engine is the Surrogate Method, which relaxes integer constraints to solve a continuous surrogate problem using stochastic approximation, then maps the solution back to feasible discrete states. To demonstrate robustness, the authors test this algorithm under two distinct traffic assignment models: a static deterministic User Equilibrium model and a dynamic platoon simulation model that accounts for queue propagation and spill-back effects. The objective function in both cases is the minimization of total weighted delay across the network. Numerical experiments were conducted on a real-world test network representing a district in Rome, comprising 51 centroids, 300 nodes, 870 links, and 70 signalized junctions. When using the User Equilibrium model, the Surrogate Method proved superior to PSO, consistently escaping local minima to achieve solutions approximately 6% better than PSO and 2% better than GA. Although GA required fewer iterations to converge, SM provided more reliable global performance regardless of the initial starting point. In the second experiment using the platoon simulation model on a main arterial road, the algorithm effectively handled non-stationary demand and spill-back phenomena. The results confirmed that the SM is capable of handling the non-convex nature of the delay function and the complex interactions between signal timing and traffic flow dynamics. The significance of this work lies in demonstrating the effectiveness of the Surrogate Method for solving complex, non-convex traffic optimization problems. By successfully integrating signal optimization with traffic assignment, the study provides a more holistic approach to urban traffic management than traditional isolated methods. The findings suggest that SM is a robust tool for reducing travel times, fuel consumption, and emissions by optimizing signal settings in a way that accounts for driver behavior. The paper concludes that while other methods like GA may be computationally faster, the Surrogate Method offers greater reliability in finding high-quality solutions in complex network environments, making it a valuable contribution to the field of intelligent 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.

• speed management