Statistical Modeling of Exhaust Emissions from Gasoline Passenger Cars

Abstract

In this study, an instantaneous vehicle emission model was developed and validated. A group of emission regression functions with vehicle speed and VSP as variables was established using the multiple linear regression method and embedded in the instantaneous emission model to predict vehicle emissions. The inputs of the instantaneous vehicle emission model consist of the driving cycle, vehicle parameters and accessories use, all of which are used to calculate the instantaneous vehicle specific power (VSP). The simulated results of the emission model are second-by-second emission rates, emission factors and fuel consumption over the target driving cycle. The predicted emissions as well as fuel consumption of four passenger cars were very close to the tested emission data, and the prediction errors of emission factors and fuel consumptions were acceptable.

Keywords

vehicle emission / regression model / statistical analysis / driving dynamics