AAU Student Projects - visit Aalborg University's student projects portal
A master's thesis from Aalborg University
Book cover


Individual Pitch Control for Load Mitigation

Authors

;

Term

4. term

Publication year

2017

Submitted on

Pages

94

Abstract

Denne afhandling udvikler en individuel pitch-regulator (styring af bladvinkel) for at mindske laster på en vindmølle. Arbejdet omfatter en ikke-lineær matematisk model, der beskriver både den strukturelle dynamik og aerodynamikken. Den strukturelle del inkluderer tårnets frem–tilbage-bevægelse (fore-aft) samt bladenes flap- og kantvise bevægelser. Aerodynamikken modelleres med Blade Element Momentum Theory (BEMT), en standardmetode der kobler lokale kræfter på bladene med rotorens samlede moment. Den ikke-lineære model lineæriseres for at kunne designe regulering. Først udvikles en kollektiv PI-regulator (proportional–integral), som justerer alle blade samtidigt for at holde strømproduktionen stabil. Derefter implementeres en individuel PI-regulator, som justerer hvert blad for sig for at udligne laster på rotoren forårsaget af vindskæring (forskelle i vindhastighed over rotorskiven). På grund af tidsbegrænsninger blev en modelprædiktiv regulering (MPC) ikke succesfuldt implementeret.

This thesis develops an individual blade pitch controller to reduce mechanical loads on a wind turbine. The work includes a non-linear mathematical model that captures both structural dynamics and aerodynamics. The structural model covers the tower’s fore-aft motion and the blades’ flapwise and edgewise motions. Aerodynamics are modeled using Blade Element Momentum Theory (BEMT), a standard method that links local blade forces to overall rotor momentum. The non-linear model is linearized to enable controller design. First, a collective PI (proportional–integral) controller is developed to adjust all blades together and keep power output steady. Then, an individual PI controller adjusts each blade separately to smooth out rotor loads caused by wind shear (differences in wind speed across the rotor disk). Due to time constraints, a model predictive control (MPC) approach was not successfully implemented.

[This abstract was generated with the help of AI]