Long-Horizon Direct Model Predictive Control with Reduced Computational Complexity

Abstract

<p>The paper proposes a strategy that reduces the computational burden of enumeration-based direct model predictive control (MPC) with long prediction intervals. This is achieved by combining a move blocking scheme with an educated restriction of the set of candidate solutions. To demonstrate the effectiveness of the proposed algorithm, a two-level converter connected to the grid via an LCL filter serves as a case study.</p>