A smart rotor configuration with linear quadratic control of adaptive trailing edge flaps for active load alleviation

The paper proposes a smart rotor configuration where adaptive trailing edge flaps (ATEFs) are employed for active alleviation of the aerodynamic loads on the blades of the NREL 5 MW reference turbine. The flaps extend for 20% of the blade length and are controlled by a linear quadratic (LQ) algorithm based on measurements of the blade root flapwise bending moment. The control algorithm includes frequency weighting to discourage flap activity at frequencies higher than 0.5 Hz. The linear model required by the LQ algorithm is obtained from subspace system identification; periodic disturbance signals described by simple functions of the blade azimuthal position are included in the identification to avoid biases from the periodic load variations observed on a rotating blade. The LQ controller uses the same periodic disturbance signals to handle anticipation of the loads periodic component. The effects of active flap control are assessed with aeroelastic simulations of the turbine in normal operation conditions, as prescribed by the International Electrotechnical Commission standard. The turbine lifetime fatigue damage equivalent loads provide a convenient summary of the results achieved with ATEF control: 10% reduction of the blade root flapwise bending moment is reported in the simplest control configuration, whereas reductions of approximately 14% are achieved by including periodic loads anticipation. The simulations also highlight impacts on the fatigue damage loads in other parts of the structure, in particular, an increase of the blade torsion moment and a reduction of the tower fore‐aft loads. Copyright © 2014 John Wiley & Sons, Ltd.     

Sizing and control of trailing edge flaps on a smart rotor for maximum power generation in low fatigue wind regimes

An extension of the spectrum of applicability of rotors with active aerodynamic devices is presented in this paper. Besides the classical purpose of load alleviation, a secondary objective is established: optimization of power capture. As a first step, wind speed regions that contribute little to fatigue damage have been identified. In these regions, the turbine energy output can be increased by deflecting the trailing edge (TE) flap in order to track the maximum power coefficient as a function of local, instantaneous speed ratios. For this purpose, the TE flap configuration for maximum power generation has been using blade element momentum theory. As a first step, the operation in non‐uniform wind field conditions was analysed. Firstly, the deterministic fluctuation in local tip speed ratio due to wind shear was evaluated. The second effect is associated with time delays in adapting the rotor speed to inflow fluctuations caused by atmospheric turbulence. The increase in power generation obtained by accounting for wind shear has been demonstrated with an increase in energy production of 1%. Finally, a control logic based on inflow wind speeds has been devised, and the potential of enhanced power generation has been shown by time‐domain simulations. Copyright © 2015 John Wiley & Sons, Ltd.     

Full‐scale test of trailing edge flaps on a Vestas V27 wind turbine: active load reduction and system identification

A full‐scale test was performed on a Vestas V27 wind turbine equipped with one active 70 cm long trailing edge flap on one of its 13 m long blades. Active load reduction could be observed in spite of the limited spanwise coverage of the single active trailing edge flap. A frequency‐weighted model predictive control was tested successfully on this demonstrator turbine. An average flapwise blade root load reduction of 14% was achieved during a 38 minute test, and a reduction of 20% of the amplitude of the 1P loads was measured. A system identification test was also performed, and an identified linear model, from trailing edge flap angle to flapwise blade root moment, was derived and compared with the linear analytical model used in the model predictive control design model. Flex5 simulations run with the same model predictive control showed a good correlation between the simulations and the measurements in terms of flapwise blade root moment spectral densities, in spite of significant differences between the identified linear model and the model predictive control design model. Copyright © 2013 John Wiley & Sons, Ltd.