Individual blade pitch control strategies for spar-type floating offshore wind turbines

Abstract

The aim of this thesis is to study the dynamics of spar-type Floating Offshore Wind Turbines (FOWTs) and investigate the performance of the individual blade pitch control strategies proposed to improve dynamic response of the turbine. The improvement in response is obtained in the form of reduced structural vibrations and improved power production. Wind loads are the largest source of dynamic loads on the wind turbines and it is well known that it is possible to reduce this aerodynamic loads by actively pitching the blades to the inflowing wind. The power output of the wind turbines are also regulated by active pitch control above the rated wind speed. Therefore, it's an open problem to design intelligent controller capable of optimizing the performance by reducing the aerodynamic loads and/or improving the power production. In this thesis, three different individual blade pitch control strategies have been proposed and their performance has been evaluated using a high fidelity FOWT model. \par First, a dynamic model of the spar-type FOWT is derived using Kane's method. Kane's method of deriving equations of motion of dynamic systems has emerged recently and presents certain advantages over the traditional Euler-Lagrangian energy formulation. Using Kane's method the equations of motion are derived from the kinematics of the system and solved directly by a computer without the need for human intervention. The derived dynamic model is verified again state-of-the-art simulator FAST~\citep{jonkman2005fast}. This derived model of the wind turbine is then used to investigate the performance of the proposed control strategies. \par The first individual blade pitch control strategy proposed in this thesis is designed based on the wavelet-LQR control strategy. The wavelet-LQR control strategy provides a framework for a time-varying controller by assigning frequency band dependent gain suitable for wind turbine applications. In this chapter, the effect of wave-current interaction is also investigated. It has been shown that wave-current interaction does not have a significant effect on the dynamics of the FOWTs. The proposed wavelet-LQR individual pitch controller was excellent in mitigating aerodynamic loads on the turbine and reducing the vibrations of the turbine and the floating platform. Since the design objective of the controller was vibration control, the excellent vibration control was accompanied by small increase in power variability. However, it has been argued that the individual wind turbine power variability will be less important as an offshore wind turbine will almost certainly be situated in a large wind farm and the total output of the farm will be predominantly dictated by the spatial and temporal variations of the farm rather than the variability of a single turbine. Hence, this increase in individual wind turbine power variability is judged to be acceptable considering the promising reduction in structural loads. \par The second individual blade pitch controller proposed in this thesis is a dual objective controller. It is well known from literature that reducing aerodynamic loads and power variability are two competitive objectives and improving one has a deteriorating effect on the other. To address this issue, the control strategy proposed in this chapter intelligently combines a simple PI (proportional-integral) controller with a low-authority LQ (Linear Quadratic) regulator. The results presented in this chapter show that the proposed controller is better than the existing controller in literature in both vibration control and power regulation. However, the benefits obtained by the above controllers comes at a cost of increased pitch actuation. \par While vibration control and improved power tracking was achieved with increased pitch actuation, it is also known that one of the largest contributors to turbine downtime is failure of pitch systems. As future turbines are continuously being up-scaled, controllers that increase pitch actuation will only compound this problem, unless, improved and reliable pitch systems are developed. To this end, the final individual pitch control strategy proposed in this thesis is developed under the optimal framework of Non-linear Model Predictive Control (NMPC). The preview of the inflow wind field required by the controller is assumed to be available from LIDAR measurements. The aim of the controller is to minimize blade pitch actuation while maintaining/improving tracking of rated rotor speed and mitigating structural vibrations of the FOWT. The existing baseline controller is designed solely to maintain rated rotor speed and hence pitch actuation is low and scope of further reduction is limited. However, the NMPC offers and optimal control framework and accounts for the future disturbance. This opens up the possibility of further reduction in blade pitch actuation without deteriorating the performance of the FOWT which is investigated in this chapter.