Multiple model predictive control with undisturbed switch for wind turbines system

The variable-speed, variable-pitch speed wind turbines are widely used in the wind-energy system. However, the fluctuation of power generation is serious because of the erratic speed of wind and the nonlinearity of the energy conversion systems (ECS). How to keep the ECS work well during the whole operating region is a challenge issue. In this paper, a multiple model predictive control (MMPC) using the state estimation is present to solve this issue. The whole operating region is divided into several sub-regions according to the wind's speed, in which several linear models and corresponding MPCs are constructed. The state variables of each local model are estimated by using the Kalman filter. A new switching criterion without perturbation is present to keep the MMPC works well at the different operating region switching. Considering the operating constraints, a quadratic programming procedure is used to calculate the optimal value of MMPC. Comparing the conventional PI controller, the proposed control achieves a good performance and reduces the fluctuation of power generation. Simulation results show the excellent performance of the present method.