Abstract
In the last decade, the continuous development of turbocharged diesel engines with increased power density has resulted in an impressive success story for light duty diesel vehicles. Upcoming, fundamental changes in the homologation process regarding both pollutant and CO2-emissions necessitate however an extensive, but efficient use of sophisticated aftertreatment systems in a wide range of operating conditions. Against this background, non-conventional combustion modes with multiple torque producing injections, such as heating modes for DPF regeneration or fast SCR light-off or rich engine operation for purging of the NOx storage catalyst (NSC), are key components requiring advanced control strategies.
This paper discusses a novel fuel path control strategy combining model-based feed forward control with cylinder pressure based feedback control. The control strategy adapts the injection pattern (number of injections) as well as the injection quantities depending on the targets and air path boundary conditions. The model for feed forward control is obtained by modeling different energy flows of the internal combustion engine, such as indicated high pressure work using a design of experiments (DoE) approach. Closed loop combustion control is achieved by deriving suitable combustion control parameters, such as the centroid of heat release. These parameters are calculated in real-time by means of a single zone combustion model.
The control strategy is implemented on a rapid-control-prototyping system and validated in transient and stationary operation. Achieved results are promising showing a reliable and constant torque behaviour in different combustion modes under varying ambient conditions with a considerable reduction of the calibration effort (~40%). The model based approach allows shifting rich events to more transient operation resulting in a reduction of the fuel consumption penalty up to 5 %, depending on application, test cycle and refinement of base calibration.
KEYWORDS – Fuel path, Closed-loop combustion control, Non-conventional combustion modes, Model-based control