An Application of Reference Governor to a Diesel Engine Air Path System: Implementation of a Multi-Variable Reference Modification Algorithm

This paper considers an application of a reference governor (RG) to the air path system of an automotive diesel engine. The air path system can be represented as a multi-input-multi-output (MIMO) nonlinear system which has hardware interaction between fresh air and exhaust gas paths resulting from a turbocharger and an exhaust gas recirculation (EGR). The system has several constraints which should be considered in control design, such as maximal boost pressure, turbine speed and EGR min-max constraints. In order to explicitly deal with the constraints, we employ an online RG to modify boost pressure and the EGR rate references so that their predicted values ensure the constraints. Using a prediction model obtained via system identification the RG algorithm is established based on a planar search combined with the gradient descent method and a bisectional search. An experimental result using a real engine is shown to demonstrate the effectiveness of the present method.