Physics-based modeling and estimation of exhaust manifold filling dynamics on a diesel engine equipped with flexible intake valve actuation.

摘要

Advanced combustion strategies, such as Premixed Charge Compression Ignition (PCCI), have been identified as one method to achieve future emissions and fuel economy regulations for diesel engines. These advanced combustion strategies require precise inputs to the combustion process to maintain stable combustion, maximize fuel efficiency, and minimize emissions. Given that the inputs to the combustion process can be considered outputs of the gas exchange process, an accurate understanding of the dynamics of the gas exchange process is critical to accurate control of the combustion process.;The work presented in this thesis focuses on modeling for estimation and the estimation of exhaust gas recirculation (EGR) flow. In the process of developing a candidate estimation strategy for the EGR flow, extensive efforts were focused on the development of a physically-based, control-oriented, gas exchange model; specifically, the development of an exhaust enthalpy model for application in the gas exchange model. The exhaust enthalpy model is utilized to determine the temperature of gas flow out of the cylinder. This is accomplished by first determining the conditions in the cylinder when the intake valve close (IVC) then assuming a polytropic compression process to top dead center (TDC). To maintain simplicity in the model, fuel injection is neglected, and it is assumed that combustion begins at TDC in a constant pressure process until the end of combustion (EOC). Upon completion of combustion, the expansion process is modeled as a polytropic expansion process until the exhaust valves open (EVO). The proposed model for exhaust gas enthalpy was extensively experimentally validated against 193 data points and is shown to accurately predict the exhaust gas temperature, generally well within 10% error.;Ultimately, the gas exchange model, including the exhaust gas enthalpy model, was utilized in the development of an estimation strategy for EGR flow. The estimation strategy utilizes the exhaust manifold pressure state equation and feedback from the exhaust manifold pressure sensor, as well as models of the EGR flow, charge flow, and turbine flow to estimate the EGR flow. Preliminary validation of the estimator has been accomplished utilizing both the gas exchange model and GT-POWER simulations, with the results demonstrating that with accurate models for charge flow and turbine flow, the estimator can accurately predict the EGR flow.