A High Fidelity Lumped-Parameter Engine Model for Powertrain Control Design and Validation

Abstract

The paper presents a lumped-parameter modeling approach for the characterization of internal combustion engine systems dynamics, suitable for the design, analysis and calibration of powertrain control. Through an analysis of the deficiencies of traditional filling and emptying, crank angle resolved models, a novel approach is presented to calibrate the distributed effects of charge heating, backflow, and wave propagation dynamics within the context of a zero-dimensional (lumped) model. The resulting model is capable of predicting the cycle averaged values of volumetric efficiency and torque, as well as the crank angle resolved values of cylinder pressure and indicated torque in steady state and transient conditions within accuracy comparable to a 1D gas dynamic model, while maintaining a significantly reduced computational effort. The approach is applied to a 4-cylinder spark ignition engine with dual independent variable valve timing, and validated using simulation and experimental data. Validation results allow for an evaluation of the model’s high accuracy and low computation effort.