Dynamic Validation and Sensitivity Analysis of a NOx Estimation Model Based on In-Cylinder Pressure Measurement

The incoming RDE regulation and the on-board diagnostics -OBD- pushes the research activity towards the set-up of a more and more efficient after treatment system. Nowadays, the most common after treatment system for NOx reduction is the selective catalytic reduction -SCR- . This system requires as an input the value of engine out NOx emission -raw- in order to control the Urea dosing strategy. In this work, an already existing grey box NOx raw emission model based on in-cylinder pressure signal (ICPS) is validated on two standard cycles: MNEDC and WLTC using an EU6 engine at the test bench. The overall results show a maximum relative error of the integrated cumulative value of 12.8% and 17.4% for MNEDC and WLTC respectively. In particular, the instantaneous value of relative error is included in the range of ± 10% in the steady state conditions while during transient conditions is less than 20% mainly.

Finally, a sensitivity analysis is conducted in order to understand how the model “answers” to any air and fuel parameter deviation.