| In this research, nonlinear hybrid control is introduced to coordinate GDI engine, emission aftertreatment systems and transmission. Dynamic modeling, parametric identification, robust control and optimization are utilized in the automotive hybrid control systems. The objective is to achieve an overall optimal performance in terms of fuel economy, emission levels, tracking property and robustness. The models are developed from fundamental electrical, mechanical and chemical relationships. Thermodynamics theory is used to analyze the internal mechanisms of exhaust emissions. Least squares estimation is applied to the transient fuel model and accurate curve fittings are obtained.; GDI engine has the essential benefit of fuel economy due to its stratified combustion. TWC treats HC and CO emissions efficiently but it is ineffective to treat NOx in lean mode. So lean NOx trap (LNT) technique is introduced to control excessive NOx emission during the lean burn. A control strategy is developed for the effective storage and purge operations of LNT. The impact of oxygen effect on NOx purge operation is analyzed for typical operating conditions. The effects of LNT temperature on engine thermal efficiency, LNT storage time, LNT purge time and fuel economy are investigated and a modified engine control strategy is proposed.; Powertrain systems are inherently hybrid. Nonlinear control approach is necessary to represent its nonlinear, time varying and uncertain nature. An optimal control problem is introduced to coordinate all components of the GDI engine, emission aftertreatment systems and transmission. The problem is formulated as an optimal hybrid control problem and solved by nonlinear discrete dynamic programming. |
