Multi-objective optimization torque control of wind turbine based on LQG optimal control

Considering power quality and drive train load, a kind of linear quadratic Guassian optimal control strategy based on Kalman filter is proposed in this thesis, in order to reduce gearbox fatigue load while improving output power quality of wind turbines. With this optimization strategy, system noise and measurement noise are estimated through Kalman filter, adjusting dynamic optimal gain according to the system of estimation, in order to achieve optimal control of different working point. A 3MW doubly-fed wind turbines is taken as the control object in the simulation,the results indicate that, compared with the conventional pitch control strategy, the proposed pitch controller not only reduced the fluctuation of power but also reduced the torque ripple of gearbox, the fatigue load of gearbox and other key component in drive train were mitigated.