Influence of model selection on optimal control of traffic for emissions minimisation

The coupling of microscopic traffic simulation models with emission models offers a powerful tool for assessing and optimising traffic control strategies to reduce fuel consumption and vehicle emissions. Although many studies use traffic simulation for emission analysis and designing traffic control measures, most focus on calibrating a selected traffic model to replicate observed traffic flow. This raises a critical question: are the resulting optimal emission control strategies adequately designed to account for the sensitivity of traffic models in capturing vehicle dynamics and emissions? To address this issue, we compared three car-following models-the Krauss model, the Intelligent Driver Model, and the Wiedemann model-each rooted in distinct theoretical frameworks to understand traffic dynamics. We evaluated their performance in optimising road speed limits to minimise (PMx) emissions in a school case study. A school was selected as the case because children are highly vulnerable and particularly exposed to pollutants during their school commute, and their exposure can be mitigated through optimal traffic control. Our findings reveal that, even when tuned to achieve comparable levels of traffic flow, the models displayed significant differences in their objective functions for traffic control optimisation. These discrepancies stemmed from variations in fuel consumption and particulate matter (PMx) emission patterns resulting from the traffic dynamics captured by the selected traffic model. At a macroscopic level (e.g. average speed, flow, and density), the models exhibited minimal differences. However, at a microscopic level (e.g. acceleration, deceleration rates, and deviations from the mean), pronounced differences became evident. These results highlight that while certain traffic control strategies appeared less effective, revisiting and critically examining the limitations of the models is essential to ensure robust and tailored solutions for emission reduction.