Predicting instantaneous engine-out NO_x emissions in a real-driving vehicle data scenario

Abstract

As type approval driving cycles were found to be insufficient to represent the real impact of emissions from on-road vehicles, it became clear that real-driving emissions must be considered to effectively address the current transportation challenges. The expected introduction of on-board emission monitoring is believed to require tools that will evaluate the on-road emissions conformity of vehicles over their lifetime. In this regard, the significant amount of data made available from modern vehicles opens for improved emissions prediction and diagnostics opportunities. This article explores therefore the effectiveness of predicting instantaneous engine-out [Formula: see text] emissions given in-service vehicle data alone. To this end, a 9 months daily vehicle usage database was used to calibrate and evaluate a semi-empirical [Formula: see text] emission model and the results showed good agreement with measurements provided by the production sensor. This approach offers a tool for simulating real-driving engine-out [Formula: see text] emissions that could help to comply with emission standards by feeding and updating advanced emissions control strategies.