Model Predictive Control of DFIG-Based Wind Turbine

This paper presents a centralized multivariable control strategy for the control of the DFIG wind turbine. Model of one-mass wind turbine with DFIG is represented by a third-order model. Model predictive control is applied in order to compensate inaccuracies and measurement noise. The optimization problem is recast as a Quadratic Program (QP) which is highly robust and efficient. In order to bring the problematic voltages as closely as possible to the desired values, multi-step optimization is introduced. The effectiveness of the proposed algorithm is demonstrated for a variable speed wind turbine.