This work demonstrates the application of the extremum seeking method for model calibration. A nonlinear, first-principle engine model is simplified for implementation in real-time control. A semi-empirical correlation is used for estimation of the wall heat transfer in the simplified engine model. The heat transfer coefficient is tuned using the extremum seeking approach to minimize the error between measurement and modeled cylinder pressure. Engine steady state measurement results are used for the demonstration and validation of the proposed technique.
