Nonlinear model predictive control of an internal combustion engine exposed to measured disturbances

• Nonlinear model predictive speed control applied to a gas engine system.
• Derivation of a control oriented model, incl. assumptions on low level controllers.
• Stability bounds are derived for the NMPC with a short prediction horizon.
• Verification of functionality by experimental tests for different operation points.

This work presents the design procedure of a speed controller for a large, lean burn, natural gas engine in island mode operation. This is a disturbance rejection problem with a measured, large disturbance. The core element is a nonlinear model predictive control (NMPC) algorithm that serves as outer loop controller in a cascaded control structure and generates set-points for low level control loops. The NMPC relies on a control oriented model that includes the physics based equations, assumptions on underlying control loops and constraints given by the control requirements. It is shown how to design the running cost such that the stability of the NMPC without terminal cost and constraints can be guaranteed for the nominal system and for the perturbed system exposed to parametric uncertainties and un-modeled dynamics. The functionality of the control strategy is demonstrated in simulation and by experimental results derived at the engine-testbed.