Author: Arno Ružička, dipl.ing.
Title: Identification of wind turbine piecewise affine model
Category: diploma thesis
Date: July 16th 2009
Mentor: prof.dr.sc. Nedjeljko Perić
Piecewise affine (PWA) models represent an easy and efficient method of describing nonlinear functions. Therefore PWA models are suitable for modeling highly nonlinear systems such as wind turbines. In this thesis, PWA models were used to describe the influence of wind speed, pitch angle and rotor speed on the aerodynamic torque and thrust force. Identification was done using algorithm that exploits the combined use of clustering , linear identification and pattern recognition techniques. As the result of identification, time discrete PWARX models were obtained. Each PWARX model contains parameters of certain affine submodel and coefficients of hyperplane where the submodel is valid. Very good match between identified PWA model and nonlinear wind turbine model was shown.