0
Research Papers

Robust Multivariable Pitch Control Design for Load Reduction on Large Wind Turbines

[+] Author and Article Information
M. Geyler

Institut für Solare Energieversorgungstechnik ISET e.V., Königstor 59, D-34119 Kassel, Gemanymgeyler@iset.uni-kassel.de

P. Caselitz

Institut für Solare Energieversorgungstechnik ISET e.V., Königstor 59, D-34119 Kassel, Gemany

J. Sol. Energy Eng 130(3), 031014 (Jul 03, 2008) (12 pages) doi:10.1115/1.2931510 History: Received March 27, 2007; Revised February 11, 2008; Published July 03, 2008

Abstract

This paper deals with multivariable pitch control design for wind turbines, including load reducing control objectives. Different design approaches, including collective and cyclic pitch, and robustness aspects are discussed. A control design with decoupled controllers for collective and cyclic pitch is worked out in detail, based on the $H∞$ norm minimization approach. The control design is verified by simulations with a full nonlinear model of the wind turbine, showing the potential of multivariable pitch control to actively increase damping of the first axial tower bending mode and to reduce 1p fluctuations in blade root bending moments. Multivariable control design provides a convenient way of including additional load reducing objectives into the pitch controller of wind turbines. Fatigue loading of certain components, as tower and blades, could be reduced significantly.

<>

Figures

Figure 1

Simplified model for coupled axial structural dynamics of tower and blades

Figure 2

Response of tower top acceleration and flapwise blade root bending moments to a step in axial tower top force ΔFT=10kN at standstill, comparison of the simplified linear model with a detailed model of structural dynamics

Figure 3

Comparison of flapwise and edgewise aerodynamic moment and aerodynamic thrust force caused by sinusoidal pitch angle changes for computations based on BEM with and without dynamic stall and dynamic inflow model, vWind=12m∕s, pitch angle amplitude Δθ=1deg, at first tower bending eigenfrequency of 0.6Hz

Figure 4

Block scheme for collective pitch controller design

Figure 5

Magnitude of open loop and closed loop transfer functions and weighting functions for collective pitch control

Figure 6

Plot of pitch angle weighting function Wp,0; comparison with inverted maximum singular value of control sensitivity function K0S0 and maximum singular value of nominal plant G0

Figure 7

Poles of open loop and closed loop systems for variation in aerodynamic coefficients; wind speed 12m∕s<vwind<24m∕s

Figure 8

Block scheme for cyclic pitch controller design

Figure 9

Magnitude plot for transfer function from input disturbances Myaw∕Mtilt to outputs MFl,bend,s∕MFl,bend,c: comparison open loop and closed loop for H∞∕PI controller

Figure 10

Maximum singular values for a number of additive perturbations from nominal plant for the range 12m∕s<vwind<24m∕s, 17rpm<nrotor<23rpm, comparison with inverted maximum singular values of control sensitivity function KscSsc for H∞∕PI controller

Figure 11

Simulated time series of axial and lateral wind speeds, axial tower top accelerations, flapwise blade root bending moments, rotor speed, and pitch angles; comparison of load reducing MVC and base line PI speed controller; initial wind speed vwind,0=16m∕s, step in wind speed magnitude Δvwind=+1m∕s at time t=40s, step in wind direction Δϕwind=15deg at t=60s

Figure 12

Simulated time series of axial and lateral wind speeds, axial tower top accelerations, flapwise blade root bending moments, rotor speed and pitch angles; comparison of load reducing MVC and base line PI speed controller; initial wind speed vwind,0=20m∕s, step in wind speed magnitude Δvwind=+1m∕s at time t=40s, step in wind direction Δϕwind=15deg at t=60s

Figure 13

Time series of average wind speed, rotor speed, pitch angle, and pitch actuator torque for a simulation of turbulent wind conditions; comparison of load reducing MVC and base line PI speed controller

Figure 14

Square roots of power spectrum of axial tower foot bending moment, flapwise blade root bending moment, and pitch actuator torque for a simulation of turbulent wind conditions; comparison of load reducing MVC and base line PI speed controller

Discussions

Some tools below are only available to our subscribers or users with an online account.

Related Content

Customize your page view by dragging and repositioning the boxes below.

Related Journal Articles
Related Proceedings Articles
Related eBook Content
Topic Collections