High Frequency IMEP Estimation and Filtering for Torque Based SI Engine Control

Torque based engine control seems to be the trend for the future for powertrain management (automatic gearbox, hybrid vehicles).

Today, torque estimation is best achieved using cylinder pressure transducers. This paper proposes a method to achieve a good accuracy of the torque using Bézier curves to reconstruct the cylinder pressure peak from the low frequency embedded pressure measurements.

As is, IMEP cannot be used on a cycle to cycle basis for engine torque control, due to the very high cycle to cycle variability of SI engines. To improve the quality of the IMEP feedback data, this paper proposes a moving horizon filtering method.