(Wiley, 2021) Elso Torralba, Jorge; Ostolaza, J. Xabier; Ingeniería; Ingeniaritza; Universidad Pública de Navarra / Nafarroako Unibertsitate Publikoa
Recent developments in quantitative feedback theory (QFT) lead to feedforward design problems with both magnitude and phase constraints. In these cases, manual feedforward tuning becomes much more challenging and time consuming than the traditional prefilter shaping taking place on the Bode plot. This article presents a general procedure for the automatic synthesis of such elements. Feedforward bounds in the complex plane are expressed as constraints of a linear programming problem in which the Bode real-complex relation is implicitly considered, ensuring a stable rational solution. The methodology is successfully tested in a well-known benchmark problem.
Wind energy tendency towards larger rotors and higher hub heights reach a limit when wind turbine pitch controller bandwidth is constrained by the first tower fore-aft bending moment. While most manufacturers try to find a workaround with alternative nonstandard tower designs, this paper presents a control solution based on quantitative feedback theory (QFT) that provides effective disturbance rejection while maintaining cost-effective conventional tower design for very high hub heights. Frequency domain specifications on measured and nonmeasured variables provide the control designer the insight to make the best use of the available feedback. The new control design procedure is applied to a realistic wind turbine example of 180-m hub height, showing the technical viability of the solution and a profitable business case.
