Exhaust pressure estimation and its application to variable geometry turbine and wastegate diagnostics

Abstract

Exhaust pressure is a critical engine parameter used to calculate engine volumetric efficiency and EGR flow rate. In this paper, exhaust pressure is estimated for an internal combustion engine equipped with a variable geometry turbocharger. A coordinate transformation is applied to generate a turbine map for estimation of the exhaust pressure. This estimation can be used to replace an expensive pressure sensor for cost saving. On the other hand, for internal combustion engines that have already installed exhaust pressure sensors, this estimation can be used to generate residual signals for model-based diagnostics. Based on the residual signals, two diagnostic methods are proposed: one based on cumulative sum algorithms and the other based on pattern recognition and neural networks. The algorithms are able to detect and isolate different failure modes for a turbocharger system.