Biologically-Inspired Neural Network for Coordinated Urban Traffic Control: Parameter Determination and Stability Analysis

Traffic congestions are a major concern for big cities around the world due to its multifaceted negative impacts. A cost-effective solution to reduce vehicle travel times and prevent traffic congestions is traffic signal control. In this work, we investigate a biologically-inspired neural network, which, in contrast to other approaches, is able to continuously monitor the system state and make decisions. An extension of a previous model is proposed, establishing a multiagent system and allowing the coordinated control of multiple intersections. Methods for parameter determination and stability analysis are also proposed. Finally, the model performance for different sets of parameters and vehicle demands is evaluated with a simulator of urban mobility and compared to a conventional cycle-based control method.