Traffic Signal Optimization: Minimizing Travel Time and Fuel Consumption

Abstract

This work integrates a multi-objective evolutionary algorithm
with the multi-agent transport simulator MATSim and the comprehensive
modal emission model simulator CMEM to analyze the evolutionary opti-
mization of traffic signals minimizing travel time and fuel consumption on
a real-world large scenario. We simulate the movement of 20.000 vehicles
on the transport network of a 5×8Km area of Quito including 70 sig-
nal lights. Our aim is to clarify the nature and the extent of the conflict
between these objectives. We also compare with a single-objective opti-
mization algorithm where only travel time is optimized and evaluate the
impact of the signals settings on gas emissions.