Including engine data for energy and pollutants assessment into the vehicle specific power methodology

Abstract

Vehicle emission models are relevant for evaluating the performance of vehicle technologies and help in the definition of environmental policies. This paper presents an improved emissions modelling approach (named VSP+M) by combining the vehicle specific power (VSP) with load-regime engine maps for each VSP mode. The new modelling enabled to link tailpipe emissions to vehicle and engine operating conditions, obtained from real driving emission (RDE) tests and on-board diagnosis (OBD) data. The parameters for the sizing of engine maps were optimised by means of Pareto frontiers to solve the trade-off between the minimisation of RMSE and emission factor errors in urban sections and total RDE trips. The CO2 emission factors errors were reduced up to 63% and 45% for urban and RDE sections, respectively. The NOx emission factor errors were reduced up to 15%, maintaining the same RMSE levels. Optimal engine maps sizing was found for every tested vehicle and for each engine type to be applied in other vehicles. This study demonstrates the potential to address declinations of the conventional VSP model based on engine operation or proxies to those variables by using the proposed approach.