Using fully variable valve control for cylinder-individual reference tracking with emission-optimal internal exhaust gas recirculation

A fully variable valve train significantly increases the degree of freedom control internal combustion engines. Cylinder deactivation, thermal management, alternative strategies, and minimized pumping losses are just a few examples enabled by freely adaptable intake exhaust timings. This paper presents method to achieve accurate tracking load trajectories under stoichiometric conditions. feedback controller is designed with mixed-sensitivity H∞ synthesis method. The underlying system plant modeled combination mean-value model cylinder-internal processes neural network map correlation between timings cylinder charge. All experiments conducted on test bench spark-ignited engine equipped an internally developed called FlexWork. With this method, mean absolute error 0.07 bar in indicated pressure 0.009 air–fuel equivalence ratio achieved for reference trajectory. Furthermore, cost function dependent online optimization gas recirculation without affecting performance stoichiometry. Depending parametrization function, nitrogen oxide or hydrocarbon pollutants can be reduced up 46% 17%, respectively.