Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

A Micro-Simulation Study of the Generalized Proportional Allocation Traffic Signal Control

Gustav Nilsson, Giacomo Como

TL;DR

The study evaluates a decentralized Generalized Proportional Allocation (GPA) traffic signal controller by discretizing it for micro-simulation and comparing it to MaxPressure in both a synthetic Manhattan-like grid and the LuST Luxembourg scenario. GPA with full or shortened cycles is shown to improve performance over static timing at low demand and remains competitive with MaxPressure at low-to-moderate demand, while MaxPressure can outperform GPA under high demand. The approach requires only local queue-length information, avoiding reliance on downstream routing data, and is demonstrated in realistic settings using SUMO and the LuST network with CVXPY-based optimization. The findings highlight a practical, distributed control strategy with tunable cycle-length behavior and potential for auto-tuning to further enhance real-world deployment.

Abstract

In this paper, we study the problem of determining phase activations for signalized junctions by utilizing feedback, more specifically, by measure the queue-lengths on the incoming lanes to each junction. The controller we are investigating is the Generalized Proportional Allocation (GPA) controller, which has previously been shown to have desired stability and throughput properties in a continuous averaged dynamical model for queueing networks. In this paper, we provide and implement two discretized versions of the GPA controller in the SUMO micro simulator. We also compare the GPA controllers with the MaxPressure controller, a controller that requires more information than the GPA, in an artificial Manhattan-like grid. To show that the GPA controller is easy to implement in a real scenario, we also implement it in a previously published realistic traffic scenario for the city of Luxembourg and compare its performance with the static controller provided with the scenario. The simulations show that the GPA performs better than a static controller for the Luxembourg scenario, and better than the MaxPressure pressure controller in the Manhattan-grid when the demands are low.

A Micro-Simulation Study of the Generalized Proportional Allocation Traffic Signal Control

TL;DR

The study evaluates a decentralized Generalized Proportional Allocation (GPA) traffic signal controller by discretizing it for micro-simulation and comparing it to MaxPressure in both a synthetic Manhattan-like grid and the LuST Luxembourg scenario. GPA with full or shortened cycles is shown to improve performance over static timing at low demand and remains competitive with MaxPressure at low-to-moderate demand, while MaxPressure can outperform GPA under high demand. The approach requires only local queue-length information, avoiding reliance on downstream routing data, and is demonstrated in realistic settings using SUMO and the LuST network with CVXPY-based optimization. The findings highlight a practical, distributed control strategy with tunable cycle-length behavior and potential for auto-tuning to further enhance real-world deployment.

Abstract

In this paper, we study the problem of determining phase activations for signalized junctions by utilizing feedback, more specifically, by measure the queue-lengths on the incoming lanes to each junction. The controller we are investigating is the Generalized Proportional Allocation (GPA) controller, which has previously been shown to have desired stability and throughput properties in a continuous averaged dynamical model for queueing networks. In this paper, we provide and implement two discretized versions of the GPA controller in the SUMO micro simulator. We also compare the GPA controllers with the MaxPressure controller, a controller that requires more information than the GPA, in an artificial Manhattan-like grid. To show that the GPA controller is easy to implement in a real scenario, we also implement it in a previously published realistic traffic scenario for the city of Luxembourg and compare its performance with the static controller provided with the scenario. The simulations show that the GPA performs better than a static controller for the Luxembourg scenario, and better than the MaxPressure pressure controller in the Manhattan-grid when the demands are low.

Paper Structure

This paper contains 14 sections, 13 equations, 18 figures, 4 tables, 3 algorithms.

Table of Contents

  1. Introduction
  2. Notation
  3. Model and Problem Formulation
  4. Feedback Controllers
  5. GPA with Full Clearance Cycles
  6. GPA with Shorted Cycles
  7. MaxPressure
  8. Comparison Between GPA and MaxPressure
  9. Simulation setting
  10. GPA Results
  11. MaxPressure Results
  12. Summery of the Comparison
  13. LuST scenario
  14. Conclusions

Figures (18)

  • Figure 1: The phases for the junction in Example \ref{['ex:phasesandprogram']}. This junction has four incoming lanes and two phases, $p_1 = \{l_1, l_3\}$ and $p_2 = \{l_2, l_4\}$. Hence there is no specific lane left-turning left.
  • Figure 2: Example of a signal program for the junction in Example \ref{['ex:phasesandprogram']}. In this example the signal program is $\mathcal{T} = \{ (p_1, 25), (p_1', 30), (p_2, 55), (p_2', 60)\}$.
  • Figure 3: How the traffic volumes evolve in time together with the cycle times for the system in Example \ref{['ex:unstableunboundedcycle']}. We can observe that the cycle length increases for each cycle.
  • Figure 4: The Manhattan-like network used in the comparison between GPA and MaxPressure.
  • Figure 5: The four different types of junctions present in the Manhattan grid, together with theirs phases.
  • ...and 13 more figures

Theorems & Definitions (3)

  • Example 1
  • Remark 1
  • Example 2