Experimental Study and Model Predictive Control of a Lean-Burn Gasoline Engine Coupled With a Passive Selective Catalytic Reduction System

Lean-burn gasoline engines have demonstrated 10-20% engine efficiency gain over stoichiometric engines and are widely considered as a promising technology for meeting the 54.5 miles-per-gallon (mpg) corporate average fuel economy standard by 2025. Nevertheless, nitrogen oxides (NO_x) emissions control for lean-burn gasoline for meeting the stringent Environmental Protection Agency tier 3 emission standards has been one of the main challenges toward the commercialization of highly efficient lean-burn gasoline engines in the United States. Passive selective catalytic reduction (SCR) systems, which consist of a three-way catalyst (TWC) and SCR, have demonstrated great potentials of effectively reducing NO_x emissions for lean gasoline engines at low cost. However, passive SCR operation may cause significant fuel penalty since rich engine combustion is required for ammonia generation. The purpose of this study is to develop a model-predictive control (MPC) scheme for a lean-burn gasoline engine coupled with a passive SCR system to minimize the total equivalent fuel penalty associated with passive SCR operation while satisfying stringent NO_x and ammonia (NH₃) emissions requirements. Simulation results demonstrate that the MPC approach can reduce the fuel penalty by 43.9% in a simulated US06 cycle and 28.0% in a simulated urban dynamometer driving schedule (UDDS) cycle, respectively, compared to the baseline control, while achieving over 97% DeNO_x efficiency and less than 15 ppm tailpipe ammonia slip. The proposed MPC controller can potentially enable highly efficient lean-burn gasoline engines while meeting the stringent Environmental Protection Agency tier 3 emission standards.