Michael Hopka

Identification of a Mean Value Model of a Modern Diesel Engine for Control Design

The modern diesel engine with its multiple subsystem interactions is a very complex plant. Especially fascinating are the interactions of the VGT and EGR control actions in their cumulative effect on the overall air loop dynamics. To circumvent the problem of manpower intensive engine mapping for determining the best operating control map, a generic mean value model of a diesel engine, based on physical principles and empirical definitions where necessary, was developed. In this paper we present the identification of this model with respect to a Fiat 2.4L JTD 166. It will be shown that there is an essential set of parameters that must be identified in order to model processes that are too complicated for physics based modeling. The parameter set is utilized in simplified empirical relationships to model complex engine phenomena such as combustion and orifice flow through the EGR and VGT, facilitating model portability. Results from matching these empirical relationships to the engine of interest and model predictions of air loop system response to EGR and VGT control inputs show good agreement with actual engine performance and are presented in this paper.Copyright