A Simple, Naive Agent-based Model for the Optimization of a System of Traffic Lights: Insights from an Exploratory Experiment

In this paper, two basic Distributed Coordination of Exploration and Exploitation algorithms, Static Estimation Optimistic K1 and Static Estimation Optimistic K2, were implemented and tested against standard isolated traffic actuated signals through simulation on a microscopic traffic simulator. The simple pilot exploratory experiment demonstrates how Naïve, Optimistic agent-based models were able to improve the performance of a system of traffic lights over time. Results also show the importance of a few variables on the system performance; namely, signal operation scheme, cool down time, and evaluation interval.