Stochastic model predictive control for wind turbines with doubly fed induction generators

A key issue in the control of double fed induction generator (DFIG) wind turbine systems is to achieve an expected performance in the presence of stochastic wind speeds. To address this issue, this paper establishes a stochastic model predictive control (SMPC) scheme for the DFIG wind turbine system. The salient feature of this scheme is that it explicitly takes into account the uncertainties in wind speed forecasts. Using wind speed predictive distributions, the forecasted wind speeds are modeled as Gaussian disturbances. Using probabilistic constraints, the uncertainties in these disturbances are incorporated into the SMPC problem formulation. By converting the probabilistic constraints into deterministic constraints, the formulated stochastic programming problem is recast as a convex quadratic optimization problem, which can be solved very efficiently. In this way, the computed control actions handle the uncertainties associated with the wind speed forecasts, thus ensuring the optimal operation of DFIG. Simulation results validate the effectiveness of the proposed scheme.