IN-CYLINDER CONDITION ESTIMATION AND CONTROL APPLICATIONS ON DIESEL ENGINE COMBUSTION

Abstract

Advanced combustion modes offer promising solutions for both emission reduction and efficiency improvement. The lower local equivalence ratio and lower peak temperature characterized by the advanced combustion mode significantly reduce the generation of the engine-out emissions (especially the soot and NOx). Although the advanced combustion mode enjoys extra-low emissions, some technical challenges prevent it from being widely applied in real practice. Combustion phasing control as auto-ignition and narrow load range are two main challenges to be addressed. The estimation and control techniques for Diesel engine targeting these two challenges are presented in four papers in this thesis. Accessing to the in-cylinder conditions is essential for a more detailed combustion estimation and further combustion control. Paper 1 and Paper 2 (Chapter II and Chapter III) introduce methods of estimating two critical in-cylinder conditions, the in-cylinder temperature and oxygen concentration. The system dynamic models are derived and the Extended Kalman filter (EKF) and smooth variable structure filter (SVSF) are utilized for the in-cylinder temperature and in-cylinder oxygen concentration estimation, respectively. The method of coordinated control for the intake conditions and the combustion process aiming at a fast and accurate combustion process response is proposed in paper 3 (Chapter IV). Disturbance rejection control in conjunction with sliding mode method is proposed to control the air- and fuel-path loop simultaneously. As an indicator to show the combustion quality and to avoid significant incomplete combustion, the unburned fuel is estimated in paper 4 (Chapter V) based on the oxygen concentration. Three filters are designed to estimate the trapped unburned fuel and their robustness against modeling errors are analyzed and compared theoretically.