Robust steady‐state target calculation for model predictive control

Abstract

Abstract In practice, model predictive control (MPC) algorithms are typically embedded within a multilevel hierarchy of control functions. The MPC algorithm itself is usually implemented in two pieces: a steady‐state target calculation followed by a dynamic optimization. A new formulation of the steady‐state target calculation is presented that explicitly accounts for model uncertainty. When model uncertainty is incorporated, the linear program associated with the steady‐state target calculation can be recast as a second‐order cone program. This article shows how primal‐dual interior‐point methods can take advantage of the resulting structure. Simulation examples illustrate the effect of uncertainty on the steady‐state target calculation and demonstrate the advantages of interior‐point methods.