Real-Time Embedded Models for Simulation and Control of Clean and Fuel-Efficient Heavy-Duty Diesel Engines 2020-01-0257
This paper presents a framework for modeling a modern diesel engine and its aftertreatment system which are intended to be used for real-time implementation as a virtual engine and in a model-based control architecture to predict critical variables such as fuel consumption and tailpipe emissions. The models are specifically able to capture the impact of critical control variables such as the Exhaust Gas Recirculation (EGR) valve position and fuel injection timing, as well as operating conditions of speed and torque, on the engine airpath variables and emissions during transient driving conditions. To enable real-time computation of the models, a minimal realization of the nonlinear airpath model is presented and it is coupled with a cycle averaged NOx emissions predictor to estimate feed gas NOx emissions. Then, the feedgas enthalpy is used to calculate the thermal behavior of the aftertreatment system required for prediction of tailpipe emissions. The complete engine and aftertreatment system models were implemented on a rapid prototyping controller and experimentally validated over steady state and transient test cycles. Results show the performance of the reduced order model was comparable to that of the full state model in predicting the transient behavior of engine airpath dynamics and NOX emissions. With an integrated torque controller equipped with a Smith predictor that cancels communication delays, the developed models present a complete setup which can be used both as a virtual engine setup or serve in engine simulation environment or in a mode-based controller design.
Citation: Duraiarasan, S., Salehi, R., Wang, F., Stefanopoulou, A. et al., "Real-Time Embedded Models for Simulation and Control of Clean and Fuel-Efficient Heavy-Duty Diesel Engines," SAE Technical Paper 2020-01-0257, 2020, https://doi.org/10.4271/2020-01-0257. Download Citation
Saravanan Duraiarasan, Rasoul Salehi, Fucong Wang, Anna Stefanopoulou, Marc Allain, Siddharth Mahesh
University of Michigan, Daimler Trucks North America