Wind turbine system identification and individual pitch control

This paper addresses system identification and control for the wind turbine operating in region 3, where the wind is excessive and blades must pitch to regulate the rotor speed and alleviate the bending moments on the blades. This region is critical for safe operation of the wind turbine. First, this paper develops linear models for controller design of the individual and collective pitch control loop by performing system identification on direct data, instead of relying on the first principles modelling. A Linear Quadratic Gaussian regulator with integral (LQGi) action is designed to control the rotor speed at its rated speed, and another multivariable LQGi individual pitch controller is used to reduce the first harmonic of the bending moment in the blades. Furthermore, an internal model based controller is an add-on to theses existing LQGi controllers to suppress synchronous harmonics directly in the d-q coordinates. Simulation experiments on FAST demonstrates the presented identification and control methods.

I. Introduction

The rising awareness of the need for environmental protection has made renewable energy an important issue worldwide. The use of natural resources, such as sun light, tides, waves, and wind provides clean and natural energy. Therefore, renewable energies is now considered to be a favorable way to combat the energy crisis. Among these renewable sources, wind energy is particularly attractive because of its high efficiency and low price. According to global wind report annual market update 2014 [1], published by Global Wind Energy Council (GWEC), there is a total installed wind capacity of 369,597 MW all over the world in 2014.