Cylinder-by-Cylinder Engine Models Vs Mean Value Engine Models for use in Powertrain Control Applications

An investigation of the need of engine model complexity for use in powertrain control applications is presented in this paper. The engine studied is an SI-engine, but the analysis methods could easily be adapted to a CI-engine. The different engine models investigated are a cylinder-by-cylinder engine model and a mean value engine model. The way to evaluate the engine models is to compare the dynamical behavior and how the engine affects the driveline. The analysis is made by studying the engine in the frequency and time domain. The investigation shows that the mean value engine model is sufficient for use in powertrain simulations and for powertrain control design. The dynamical behavior for the two models coincide, and the combustion pulses are well damped in the driveline (powertrain without engine). A less complex structure is preferable when designing a control system. However, to study the effects of backlash it can not be concluded which degree of engine model complexity to use, since it depends on the time spent in the backlash as well as the control strategy.