Gain-Scheduled l1-Optimal Control of Variable-Speed-Variable-Pitch Wind Turbines

The fast-growing technology of large scale wind turbines demands control systems capable of enhancing both the efficiency of capturing wind power, and the useful life of the turbines themselves. l(1)-Optimal control is an approach to deal with persistent exogenous disturbances which have bounded magnitude (l(infinity)-norm) such as realistic wind disturbances and turbulence profiles. In this brief, we develop an efficient method to compute the l(1)-norm of a system. As the control synthesis problem is nonconvex, we use the proposed method to design the optimal output feedback controllers for a linear model of a wind turbine at different operating points using genetic algorithm optimization. The locally optimized controllers are interpolated using a gain-scheduled technique with guaranteed stability. The controller is tested with comprehensive simulation studies on a 5 MW wind turbine using fatigue, aerodynamics, structures, and turbulence (FAST) software. The proposed controller is compared with a well-tuned proportional-integral (PI) controller. The results show improved power quality, and decrease in the fluctuations of generator torque and rotor speed.