查看文章

This study explores the potential range of effects of driving behaviour modified by automated vehicles (AV) on Greenhouse Gas (GHG) emissions using traffic microsimulation and emissions modeling. The driving behavior parameters of a traffic simulation package, most relevant to AV operation, are tested within the ranges deemed to be representative of potential AV operations. The impact of each driving behavior parameter with respect to GHG emissions and traffic performance is assessed through one-at-a-time (OAT) and statistical sensitivity analysis approaches. The OAT approach involves incrementing each parameter independently to determine its effects. The sensitivity analysis involves calculating the Pearson, Spearman and Partial Correlation (PCC) Coefficients based on a Monte Carlo simulation of 400 combinations of the parameters. Based on the OAT approach and PCC statistic, the authors observe that headway time (CC1), oscillation acceleration (CC7), and standstill distance (CC0) are the most influential driving behavior parameters for GHG emissions. When considering interaction effects through the Pearson and Spearman correlation coefficients, the authors observe that headway time (CC1), negative/positive following threshold (CC4/CC5), and oscillation acceleration (CC7) are the most influential parameters on GHG emissions. This study also concludes that driving settings of AVs alone do not have a significant impact on GHG emission reductions, but when including the effects of vehicle powertrain technology, there is potential for up to 24% in GHG emission reduction.