No abstract

Further improvements in the cost-effectiveness of wind turbines drives designers towards larger, lighter, more flexiblestructures in which more intelligent control systems play an important part in actively reducing the applied structural loads, avoiding the need for wind turbines to simply withstand the full force of the applied loads through the use of stronger, heavier and therefore more expensive structures. Controller research within the UPWIND project has been aimed at further developing such control strategies and ensuring that new, often larger and innovative turbines can be designed to use these techniques from the start. For this to be possible, it is important to build up full confidence in the effectiveness and the reliability of these strategies in all situations. To this end, the work reported in this paper covers several different aspects: full-scalefield testing to build confidence in the effectiveness of advanced control strategies; further development of advanced control strategies to prevent unwelcome side effects in any of the load cases that have to be considered during the design; the possibility of blades employing dual-pitch control; development of load estimation techniques that can reduce reliance on additional sensors that would otherwise be required; investigating the potential of light detection and ranging assisted feed-forward pitch control to mitigate extreme and fatigue loads; using system identification methods to improve controller tuning. The detailed results of the work presented in this paper are available in the published reports of the Control Systems work package of the UPWIND project. These reports also cover other results of the work package, which are not reported here, such as control during network faults such as voltage dips, voltage control at the point of connection to the network and gradual cut-out of wind turbines to improve output predictability in high winds. A summary report is also available

No abstract

The importance of the controller in determining the design loads of a wind turbine has been recognised for many years. This paper will discuss this topic from the following perspectives:Approaches to the design and tuning of closed loop controllers: fatigue loading can be reduced by using advanced control methods, sometimes dramatically. Classical design methods currently predominate, although there may also be a role for other approaches such as model-based methods.Network faults: The response of wind turbines to network faults such as voltage dips are of increasing concern for turbine designers and network operators, and modifications to the control system may be required in order to minimise the probability of turbine shut-down following a short-duration network fault, as well as to minimise the loading consequences of such events. Advanced simulation tools capable of dealing with these events are required so that the most appropriate strategies can be devised.Interactions between closed loop and supervisory control: supervisory control events such as shut-downs can be a source of design-driving loads, although these can often be mitigated by careful consideration of the details of the shut-down control and how it may interact with the closed loop control. Also the advanced closed-loop methods may themselves have consequences for extreme loads, sometimes requiring careful consideration of shutdown strategies. The distinction between controlled shutdowns and safety system shutdowns is very important to consider.

We analyze the performance of different control schemes when applied to the regulation problem of a variable-speed representative wind turbine. In particular, we formulate and compare a wind-scheduled PID, a LQR controller and a novel adaptive non-linear model predictive controller, equipped with observers of the tower states and wind. The simulations include gusts and turbulent winds of varying intensity in nominal as well as off-design operating conditions. The experiments highlight the possible advantages of model-based non-linear control strategies.