Engine modeling and control for minimization of hydrocarbon coldstart emissions in SI engine

The need to reduce the negative effects of automotive hydrocarbon emissions on human health has drawn attention to the coldstart period. This phase consists of the first few minutes of operation of the engine and generates a large percentage of the total cumulative hydrocarbon emissions. The focus of this dissertation is on developing strategies to reduce emissions without adding extra hardware components to the engine. First, parameters are identified for an already existing engine model. A new model for fuel dynamics is developed and presented that allows to better describe the fuel film formation in the engine intake port. Simplified control oriented models are also developed for exhaust temperature and raw hydrocarbon emissions. The complete engine model fits well experimental data.; In the second part of this study, controllers were developed to reduce emissions. The first one is based on Dynamic Surface Control. The controller is designed to track simultaneously the profiles of two engine variables: desired raw hydrocarbon emissions and desired catalyst temperature. The trajectories can be set to achieve fast catalyst lightoff and low hydrocarbon emissions. These two objectives are in conflict with each other during coldstart. To achieve more flexibility, a hybrid controller was derived that allows to deal with these two competing objectives. Reachability analysis tools were used to test the stability of the controller.; The second methodology of controller design was based on Dynamic Programming. The problem of minimization of hydrocarbon emissions was set up in such a way that optimality could be analyzed with different time horizons. The results allowed to explain more clearly the tradeoff of the coldstart problem. The resulting optimal control strategy can be used to define tracking profiles in the Dynamic Surface Controller developed previously. An approximation of the optimal control problem was also formulated. Multiparametric optimization techniques were used to outline an explicit control law for the problem.