Nonlinear observer design for turbocharger in a SI engine

Estimation of major turbocharger variables is essential for proper control and monitoring of a turbocharged engine. This work presents a novel algorithm to estimate turbocharger rotator speed, air temperature downstream a compressor and flow rate over the compressor using mean value models of engine subsystems. A nonlinear Luenberger observer is designed for a 1.7-lit gasoline turbocharged engine. The designed observed is shown analytically to be asymptotically stable. Performance of the designed observer is experimentally validated with the data collected from the engine. The results indicate the observer can capture major turbocharger's rotational dynamics and estimated turbocharger variables are in a good agreement with the experimental measurements.