Adaptive Model Predictive Control of Combustion in Flex-Fuel Heavy Duty Compression-Ignition Engine

Abstract

Flex-fuel engines can operate on different fuels, from fossil fuel to renewable fuel and their mixture. With the assumption that fuel species is unknown in advance, the mutative fuel properties give rise to an interesting control problem. Since the combustion phasing and ignition delay in the combustion process are intimately coupled, the fuel injection system and air system need to be combined for performance. In this work, an adaptive Model Predictive Control (MPC) approach is proposed to control the combustion process in a multi-cylinder heavy duty compression-ignition (CI) engine. MPC is a suitable design for this multiple inputs/outputs system with actuator constraints, and adaptivity is the solvent for the unknown mutative fuel properties. The combustion timing and ignition delay are extracted from cooled in-cylinder pressure sensors and simultaneously controlled by manipulating injection timings, the intake oxygen concentration, and intake pressure using an exhaust-gas recirculation (EGR) system and a variable-geometry turbocharger (VGT). Diesel, gasoline/n-heptane mixture, and ethanol/n-heptane mixture are used in the experiments. The method is validated in fuel transitions from diesel to gasoline mixture and from gasoline mixture to ethanol mixture.