Variable
geometry turbines (VGT) are of particular interest to advanced
diesel powertrains for future conventional trucks, since they
can dramatically improve system transient response to sudden changes
in speed and load, characteristic of automotive applications.
VGT systems are also viewed as the key enabler for the application
of the EGR system for reduction of heavy-duty diesel emissions.
This paper applies an artificial neural network methodology to
VGT modeling in order to enable representation of the VGT characteristics
for any blade (nozzle) position. Following validation of the ANN
model of the baseline, fixed geometry turbine, the VGT model is
integrated with the diesel engine system. The latter is linked
to the driveline and the vehicle dynamics module to form a complete,
high-fidelity vehicle simulation. A virtual Class VI vehicle is
tested during hard acceleration and over a complete highway driving
cycle in order to evaluate the effect of VGT on engine-in-vehicle
response and cycle parameters affecting fuel economy and emissions. |