Environmental Performance Measures in Optimizing Traffic Signals for Fuel and Emissions Efficiency

Designing traffic signal timings is a cost-effective method of improving traffic flows along urban arterials. Proper signal timings can significantly enhance primary mobility metrics (delay and stops), and significantly improve traffic safety and sustainability measures (Fuel Consumption ‘FC’ and Vehicular Emissions ‘VEs’). The last decade has seen a growing concern to minimize traffic sustainability measures through various transportation applications. The overarching theme of this research is to develop a framework for signal timing optimization to improve traffic sustainability measures. The research first developed an Environmental Performance Index (Env-PI) as a linear relationship between delay and stops with a variable “K” (aka stop penalty) that assigns an FC weight to each stop. In addition, the research investigated individual impacts of multiple operating conditions (e.g., vehicle type, speed) on stop-induced FC and K, in simulation environment. The results showed that various operating conditions affect the K differently. Consequently, the research then explores the compound influence of those conditions on the K value, but this time from both FC and VEs perspectives. The outcomes of such experiments indicated that the K varies significantly for various combinations of conditions and sustainability metrics, suggesting that minimizing FC does not necessarily minimize all VEs. The research subsequently developed predictive models for the FC-based Ks utilizing FCs collected from a large vehicular fleet from the field. The models’ estimates are shown to be very accurate; hence, they are used to validate the simulated K values obtained in the earlier research stages. The findings revealed that the simulated Ks strongly correlate with the field Ks; thus, simulated Ks are credible to be implemented in signal optimization practice. In the next step, the FC-PI was deployed as an objective function to optimize signals on a corridor with 13-signalized intersections. The results show that FC-PI could achieve significant FC and VEs savings. Finally, this research proposed a methodology to integrate the effect of all deceleration-acceleration events (both full and partial stops) in the Env-PI to increase its accuracy. Collectively, this research is a significant step to facilitate a novel practical approach to optimize signal timings to reduce FC and VEs