Wind turbines are fatigue critical machines for which the problem of vibrations is becoming predominant as rotor sizes keep increasing. Whereas the load reduction potential of wind turbine smart blades with trailing-edge control devices is now well-recognised, the design of load alleviation control systems which are efficient, robust and yet simple remains an open problem. The present paper evaluates the efficiency of a new proposed decoupled loop-shaping control (DLSC) strategy for the load alleviation of wind turbine blades equipped with multiple control surfaces. The efficiency of the proposed decoupling control strategy is evaluated against the well-established model predictive controller (MPC). For that purpose, an in-house wind turbine aeroelastic control (WTAC) simulator is employed to simulate the aeroelastic dynamics of the NREL 5 MW wind turbine operating in a fully turbulent environment. WTAC incorporates a wind field generator, an aerodynamic model and a finite element structural blade model. Trailing edge flaps are used as active control surfaces and are dynamically coupled to the aeroelastic wind turbine blade models. First, a state-of-the-art MPC is designed to address the load alleviation control problem. MPC is specifically chosen on account of its multi-input multi-output (MIMO) nature and constraints handling capabilities. Although high load alleviation efficiency is achieved, the MPC remains a complex timebased state controllers. Next, the proposed frequency-based loop-shaping controller is applied to the same control problem. The DLSC takes advantages of the dynamic response of blades to reformulate the MIMO load alleviation control problem as decoupled single-input single-output (SISO) control loops. Classic loop-shaping control theory is then used to design SISO controllers. The DLSC performance is compared with that of the MPC based on 1P blade root load bending moment reduction. The results of this investigation demonstrate the efficiency of the proposed DLSC strategy with respect to the MPC controller. The advantageous simplicity and insightful features of the DLSC are also highlighted.
|Number of pages||1|
|Publication status||Published - 2015|
|Event||26th International Conference on Adaptive Structures and Technologies, ICAST 2015 - Kobe, Japan|
Duration: 14 Oct 2015 → 16 Oct 2015
|Conference||26th International Conference on Adaptive Structures and Technologies, ICAST 2015|
|Period||14/10/15 → 16/10/15|