Development and Validation for Mean Value Engine Models

Engine control algorithms are among the most important factors that affect engine performance and emission. Developing control algorithms would improve engine performance, fuel consumption and emission levels. On the other hand, time and cost reduction of controller development is becoming an ever increasing demand. To meet these demands, more advanced engine models and better controller development processes are essential. Therefore, those models with good accuracy together with high calculation speed and fewest numbers of tests for calibration are most suitable. The mean value engine models are developed to meet these criteria. The governing equations for these models are simple and relatively easy to calibrate. The main purpose of this work was to simplify the equations of such a model, decrease the number of calibration tests and improve model accuracy. Simpler equations are used for the calculation of air mass flow at the throttle body and cylinder ports. To increase the accuracy of the manifold pressure calculations, two different relations are proposed and the results are compared. Also a set of equations is presented for rotational dynamics. Then the accuracy of the developed model is examined through the experimental works carried out on the engine of a locally manufactured vehicle called Samand.