Design and Simulation-Based Testing of Connected Traffic Light Guidance Systems

Abstract: The establishment of fast and reliable communication technologies, such as 5G, is enabling the evolution of a new generation of connected ADAS. This work aims to develop a traffic light advisory system, Multiple Traffic Light Advisor (MTLA), to improve driving efficiency and intersection viability, and reduce urban pollution. The developed system guides the driver on how to modify the vehicle speed to efficiently utilize the current and future states of the traffic lights ahead. Starting from a non-optimal implementation of the overall architecture, MTLA is further improved through a non-linear MPC approach. The developed system is tested in a virtual environment in IPG CarMaker and results show good performances with a high potential and space for future developments.