CI Engine Model Predictive Control With Availability Destruction Minimization

Abstract

Major interests in the automotive industry include the use of alternative fuels and reduced fuel usage to address fuel supply security concerns and regulatory requirements. The majority of previous internal combustion engine (ICE) control strategies consider only the First Law of Thermodynamics (FLT). However, FLT is not able to distinguish losses in work potential due to irreversibilities, e.g., up to 25% of fuel exergy may be lost to irreversibilities. To account for these losses, the Second Law of Thermodynamics (SLT) is applicable. The SLT is used to identify the quality of an energy source via availability since not all the energy in a particular energy source is available to produce work; therefore optimal control that includes availability may be another path toward reduced fuel use. Herein, Model Predictive Control (MPC) is developed for both FLT and SLT approaches where fuel consumption is minimized in the former and availability destruction in the latter. Additionally, both include minimization of load tracking error. The controls are evaluated in the simulation of a single cylinder naturally aspirated compression ignition engine that is fueled with either 20% biodiesel and 80% diesel blend or diesel only. Control simulations at a constant engine speed and changing load profile show that the SLT approach results in higher SLT efficiency, reduced specific fuel consumption, and decreased NOx emissions. Further, compared to use of diesel only, use of the biodiesel blend resulted in less SLT efficiency, higher specific fuel consumption, and lower NOx emissions.