Estimation and Prediction of In-Cylinder Chemical Species in a Gasoline Engine for Control Purposes

Abstract

To simultaneously reduce pollutant emissions and maximize the efficiency of spark ignition engines, several technologies have been developed that must be precisely controlled. For example, exhaust gas recirculation (EGR) is used to minimize NOx production, and running SI engines lean increases the global efficiency. Both these technologies have a deep impact not only on the air mass inside the cylinder, but also on the composition of the enclosed gas. Intake manifold pressure and temperature coupled to a combustion equation permit an estimate of the in-cylinder gas composition which is predominant for combustion process behavior. In-cylinder gas composition must be known before the injection of the fuel is performed, to control the engine performance and emissions. This paper proposes a linear Luenberger observer based on a physical model to predict in-cylinder gas composition for engine control purposes.