Modellbasierte optimale Mehrgrößenregelung und optimale Reglerparametrisierung für Luftsysteme von Pkw-Dieselmotoren

Abstract

With legislation restricting emissions and customers increasing their demands, the air system of modern diesel engines has evolved into a complex system. Several coupled actuating and controlled variables have to be considered simultaneously in mathematical modelling and control. This thesis investigates the use of a linear-quadratic regulator (LQR) as a model-based, optimal, multivariable control for a diesel engine's air system. This technique is validated by results of simulation studies and on-board tests. The challenge of determining appropriate regulator parameters is addressed by a new method, which calculates optimal regulator parameters as the solution of suitable optimal control problems.