In-Vehicle Realization of Nonlinear MPC for Gasoline Two-Stage Turbocharging Airpath Control

Innovative charging concepts, such as two-stage turbocharging for gasoline engines, cause high demands on the process control. The open-loop process is characterized by a complex, nonlinear system behavior. In addition, the requirements on the closed-loop system are challenging: fast reference tracking has to be achieved without overshoots while respecting constraints on the turbocharger speeds in order to prevent damaging of the components. Nonlinear model predictive control (NMPC) offers a high potential for this purpose. It is capable of handling coupled multiple-input systems while achieving high control quality and respecting constraints on system states. This paper presents an NMPC scheme for a two-stage turbocharged gasoline airpath, which is based on a physically driven reduced-order model formulated as a set of differential-algebraic equations. The online optimal control algorithm uses the real-time iteration scheme and is implemented on a control prototyping platform. For validation of the algorithm, it is tested based on simulations and vehicle experiments. These experiments have been conducted on a vehicle dynamometer as well as on an automotive testing track. For this purpose, a modified production vehicle is used in which the airpath concept is implemented.