Performance Simulation And Control Design For Diesel Engine Nox Emission Reduction Technologies

Fuel efficiency and emission reductions are the two consistent drivers for internal combustion engine development for both on-highway and off-road vehicles. Advanced combustion technologies are proposed for the improvement of fuel consumption and reduction of harmful gas production inside the cylinder in laboratory engines. Outside cylinder technologies and after-treatment are the alternatives for a production engine to meet the stringent emission standards. Advanced control technologies play important roles in the realization of new technologies. This research was aimed at investigating possible techniques and feasible methods of implementation to reduce diesel engine emissions to meet the more stringent Tier 4 standards. In this study, two technologies are studied for off-road diesel engine NOx emission reductions: stoichiometric combustion ignition (SCI) and lean NOx trap (LNT). The concept of the stoichiometric compression ignition (SCI) engine was investigated for implementation in a turbocharged diesel engine through co-simulation. At first, an integrated environment for 1D engine modeling with control function was proposed for a SCI performance evaluation and control implementation. The SCI engine has been evaluated by Constant Speed Load Acceptance tests under steady-state and transient conditions. For SCI implementations, basic controls have been designed including air-fuel ratio (AFR) control, torque limiting control and idle speed control. The proposed control strategies have been verified with 1D detail models in the integrated environments. Further, the Mean Value Engine Model (MVEM) is proposed for advanced model based control design. The SCI engine subsystems are modeled using an orifice constrain model for throttle, turbine, and wastegate; filling and emptying model for intake and exhaust manifolds; rotational dynamic for engine camshaft and turbocharger shift, air-charging model and exhaust properties regressed by the data from integrated simulation at different engine operating conditions. The MVEM was implemented in Matlab/Simulink for verification. Modular and system verification was conducted for steady-state and transient state consistency with the 1D detail model. The results are promising, but the whole system needs further tuning for dynamic control design. The lean-NOx trap, as an alternative after-treatment for NOx control, has been studied for generic diesel engine emission control. Based on experimental data, an improved NOx adsorption model is proposed for integrated engine control and optimization.