Fake algebraic Riccati equation applied to model predictive control for doubly fed induction generator direct power control

Abstract

Model Predictive Control (MPC) and Direct Power Control (DPC) are widely used in Doubly Fed Induction Generator (DFIG) applications. However, in the context of DPC‐based MPC for DFIG, there has been a lack of literature focusing on the design of the MPC parameters. In this way, this article introduces a novel technique for designing weighing matrices in Direct Power Model Predictive Control using the Fake Algebraic Riccati Equation (FARE). As FARE is associated with the infinite receding horizon, a small receding horizon can be designed without compromising the robustness. Furthermore, this article presents a computationally efficient linear state‐space model for DPC that, when combined with a small receding horizon FARE design, offers a controller with low computational cost and robustness. The proposed theory is validated through experimental results under normal operation, variable speed operation, and parameter variations. Additionally, the proposed controller is compared to other MPC approaches.