Past research has shown substantial reductions in the oxides of nitrogen (NOx) concentrations by using 10%-25% exhaust gas recirculation (EGR) in spark ignition (SI) engines (see Dudek and Sain, 1989). However, under high EGR levels, the engine exhibits strong cyclic dispersion in heat release which may lead to instability and unsatisfactory performance preventing commercial engines to operate with high EGR levels. A neural network (NN)-based output feedback controller is developed to reduce cyclic variation in the heat release under high levels of EGR even when the engine dynamics are unknown by using fuel as the control input. A separate control loop was designed for controlling EGR levels. The stability analysis of the closed-loop system is given and the boundedness of the control input is demonstrated by relaxing separation principle, persistency of excitation condition, certainty equivalence principle, and linear in the unknown parameter assumptions. Online training is used for the adaptive NN and no offline training phase is needed. This online learning feature and model-free approach is used to demonstrate the applicability of the controller on a different engine with minimal effort. Simulation results demonstrate that the cyclic dispersion is reduced significantly using the proposed controller when implemented on an engine model that has been validated experimentally. For a single cylinder research engine fitted with a modern four-valve head (Ricardo engine), experimental results at 15% EGR indicate that cyclic dispersion was reduced 33% by the controller, an improvement of fuel efficiency by 2%, and a 90% drop in NOx from stoichiometric operation without EGR was observed. Moreover, unburned hydrocarbons (uHC) drop by 6% due to NN control as compared to the uncontrolled scenario due to the drop in cyclic dispersion. Similar performance was observed with the controller on a different engine.

No abstract

No abstract

Research has shown substantial reductions in the oxides of nitrogen (NO x ) concentrations by using 10% to 25% exhaust gas recirculation (EGR) in spark ignition (SI) engines [1]. However under high EGR levels the engine exhibits strong cyclic dispersion in heat release which may lead to instability and unsatisfactory performance. A suite of neural network (NN)-based output feedback controllers with and without reinforcement learning is developed to control the SI engine at high levels of EGR even when the engine dynamics are unknown by using fuel as the control input. A separate control loop was designed for controlling EGR levels. The neural network controllers consists of three NN: a) A NN observer to estimate the states of the engine such as total fuel and air; b) a second NN for generating virtual input; and c) a third NN for generating actual control input. For reinforcement learning, an additional NN is used as the critic. The stability analysis of the closed loop system is given and the boundedness of all signals is ensured without separation principle. Online training is used for the adaptive NN and no offline training phase is needed.Experimental results obtained by testing the controller on a research engine indicate an 80% drop of NO x from stoichiometric levels using 10% EGR. Moreover, unburned hydrocarbons drop by 25% due to NN control as compared to the uncontrolled scenario. ( ) 1 1 m d k d < and ( ) 2 2 m d k d < . Here m d 1 and m d 2 are unknown positive scalars.

No abstract