Table of contents

The target of improving efficiency of wind kinetic energy extraction has stimulated a certain attention to wind turbine retrofitting. This kind of interventions has material and labor costs and producible energy is lost during installation. Further, the estimation of the energy enhancement is commonly provided under the hypothesis of ideal conditions that can be very different from real ones. Therefore, a precise estimation of performance improvement is fundamental. In this work, a SCADA-based method is formulated for estimating the improvement in energy production of multi-megawatt wind turbines, sited in Italy in a very complex terrain. The blades of one wind turbine in the farm have been optimized by installing vortex generators and passive flow control devices. An Artificial Neural Network (ANN) model is employed: the output is the power of the retrofitted wind turbine and the inputs are the powers of some reference nearby wind turbines. The production increase is estimated by observing how the difference between simulated and measured power output changes after the installation of the aerodynamic upgrade. The average improvement is estimated as the 3.9% of the total energy produced below rated power.

This paper investigates the use of two different liquid column dampers for vibration control of spar-type floating offshore wind turbines (FOWTs). A 16-degree-of-freedom (16-DOF) aero-hydro-servo-elastic model for the FOWT is first established using multi-body based formulation and the Euler-Lagrangian equation, taking into consideration the full coupling of the blade-drivetrain-tower-spar vibrations, a collective pitch controller and a generator controller. It is found from the simulation results that due to the coupling to the spar rigid-body motion, the eigenfrequency of the tower vibration is significantly changed, which needs to be accounted for when tuning the liquid dampers. Tuned liquid column damper (TLCD) is investigated for controlling the lightly damped (due to low aerodynamic damping) tower side-side vibration and blade edgewise vibrations. Further, a newly proposed liquid column damper, the circular liquid column damper (CLCD) is also investigated for blade edgewise vibration control. The large centrifugal acceleration from the rotating blade makes it possible to use liquid column dampers with rather small masses for effectively suppressing edgewise vibrations. By properly tuning the dampers, both types of liquid column dampers are effective in mitigating tower and blade vibrations. Performances of the dampers are compared in terms of the control efficiency and practical considerations.

This contribution investigates different noise adaptive Kalman filtering techniques with regard to their usability for wind turbine application. Since advanced model-based control schemes arise as promising alternative for standard industrial control, the necessity for robust and adaptive state estimation techniques has simultaneously emerged as an important topic. The comparison of the implemented adaptation rules shows that the master-slave adaptive filters are very flexible, numerically efficient and easy to implement. Maximum likelihood estimation based methods are more robust but show less flexibility and fewer design parameters to influence the filter performance. The simulation study shows that adaptive filters are beneficial since they solve two typical problems involved with static Kalman filter design: First, filter parameter adaptation compensates incorrect assumptions of noise statistics. Secondly, adaptation rules prevent poor filter performance for systems with time-varying statistical properties.

This paper investigates combined feedback-feedforward control of a wind turbine using a blade-mounted telescope LiDAR sensor to measure the inflow wind speed. The objective of this paper is to evaluate the controller performance with fixed telescope parameters compared to the wind speed dependent dynamically changing telescope parameters. First the telescope parameters for both cases are determined, then the robust feedback controllers are extended with an inverse-based feedforward individual pitch and trailing edge flap controller to alleviate the 1P and 2P loads of the flapwise blade root bending moments. A minor performance degradation is observed with the fixed telescope parameters in comparison to the dynamic telescope parameters.

This paper provides enhancements to normal behaviour models for monitoring major wind turbine components and a methodology to assess the monitoring system reliability based on SCADA data and decision analysis. Typically, these monitoring systems are based on fully data-driven regression of damage sensitive-parameters. Firstly, the problem of selecting suitable inputs for building a temperature model of operating main bearings is addressed, based on a sensitivity study. This shows that the dimensionality of the dataset can be greatly reduced while reaching sufficient prediction accuracy. Subsequently, performance quantities are derived from a statistical description of the prediction error and used as input to a decision analysis. Two distinct intervention policies, replacement and repair, are compared in terms of expected utility. The aim of this study is to provide a method to quantify the benefit of implementing the online system from an economic risk perspective. Under the realistic hypotheses made, the numerical example shows for instance that replacement is not convenient compared to repair.

In recent years the focus of wind energy industry is on reducing levelized cost of energy by rotor upscaling. Moreover, a current topic of interest to both industry and academia is the extension of lifetime to existing wind turbines approaching the end of initial design span. Thus, the need for load alleviation technologies integrated in the design process or for retrofit purposes is becoming more relevant. One of these is individual blade pitch control, a recurring topic in research, with known advantages and weaknesses namely the pitch actuator and bearing wear. The present work suggests such a system incorporating three independent controllers with input the root bending moments on the rotating frame. The linear system used for controller design is based on black box identification of non-linear simulations and filters are used both for the input and output. Different setups of the independent blade control scheme are applied on a 10 MW reference turbine, with a large and highly flexible rotor representative of the current industrial status, under wind conditions as defined by relevant certification standards. The investigation aims on evaluating the system's performance based on the fatigue load alleviation potential for different components as well as identifying the tradeoff for each design choice. Finally, based on basic assumptions the reductions are translated to possible life time extension for each component based on a combined operation where the new controllers are applied for a percentage of the initial 20 year lifetime.

In this paper, the turbine itself is used as an anemometer to estimate the inflow at its rotor disk. Indeed, given that any anisotropy in the wind will lead to periodic loads, by studying the machine response one can infer rotor-effective wind conditions and exploit such information for turbine and farm-level control applications. Specifically, expanding the idea of previous publications, the case of an individually pitch-controlled machine is considered herein: a linear implicit model is formulated to relate some characteristics of the wind —in the form of shears and misalignment angles— to the 1P harmonics of pitch angles and blade loads. The performance of the proposed algorithm is tested in a simulation environment, including both uniform and turbulent wind conditions.

To reduce the cost of wind energy, it is essential to reduce loads on turbine blades to increase lifetime and decrease maintenance cost. To achieve this, Individual Pitch Control (IPC) received an increasing amount of attention in recent years. In this paper, a data-driven IPC algorithm called Subspace Predictive Repetitive Control (SPRC) is used to alleviate periodic loads on a scaled 2-bladed wind turbine in turbulent wind conditions. These wind conditions are created in an open-jet wind tunnel with an active grid, enabling unique reproducible high turbulent wind conditions. Significant load reductions are achieved even under these high turbulent conditions.

Wind energy research groups from various disciplines generally use self-developed baseline wind turbine control implementations and tunings, which complicates the evaluation and comparison of new control algorithms. To solve this problem, the Delft Research Controller (DRC) provides an open, modular and fully adaptable baseline wind turbine controller to the scientific community. New control implementations can be added to the existing baseline controller, and in this way, convenient assessments of the proposed algorithms is possible. Because of the open character and modular set-up, scientists are able to collaborate and contribute in making continuous improvements to the code. The DRC is being developed in Fortran and uses the Bladed-style DISCON controller interface. The compiled controller is configured by a single control settings parameter file, and can work with any wind turbine model and simulation software using the DISCON interface. Baseline parameter files are supplied for the NREL 5-MW and DTU 10-MW reference wind turbines.

In this paper, data from a lidar-based field campaign are used to examine the effect of yaw misalignment on the shape of a wind turbine wake. Prior investigation in wind tunnel research and high-fidelity computer simulation show that the shape assumes an increasingly curled shape as the wake propagates downstream, because of the presence of two counter-rotating vortices. The shape of the wake observed in the field data diverges from predictions of wake shape, and a lidar model is simulated within a large-eddy simulation of the wind turbine in the atmospheric boundary layer to understand the discrepancy.

Turbine power and yaw set points can be adjusted across a wind farm to minimise the overall power losses and the additional fatigue loads caused by wake interactions. Detailed modelling is required to understand the complex flows in sufficient detail to allow a realistic practical control design. High-fidelity computational fluid dynamics requires enormous computational resources, so simpler engineering models are needed which capture the most important effects while running fast enough to allow sufficient testing. This paper describes a steady-state optimisation tool which has been extended to optimise all the power reduction set-points and yaw offsets simultaneously for different wind conditions. It also describes a fast time-domain simulation model which captures turbine and wake dynamic effects, so that wind farm controllers of all kinds can be tested in realistic and time-varying conditions. To demonstrate its application for controller testing, the performance of the combined power and yaw controller is tested during changing conditions of wind speed, direction and turbulence derived from measured site data. Finally, the need for validation is discussed, as many uncertainties still need to be resolved in order to obtain sufficient confidence that the potential benefits of such wind farm control schemes can be realised in practice.

In this work, a control-oriented wind farm model is validated against SCADA data from a typical on-shore wind farm, without additional instrumentation available. The comparison of model-predicted and measured power deficits due to wake impingement shows good agreement. Furthermore, the model is used to compute optimum yaw misalignments for yaw-induced wake steering, leading to an estimated 1.7% increase in annual energy production by mitigation of wake losses. Results show that wake steering based on standard SCADA data, which is usually available in operational wind farms, has promising potential for open-loop model-based wind farm control.

Wind farm control research typically relies on computationally inexpensive, surrogate models for real-time optimization. However, due to the large time delays involved, changing atmospheric conditions and tough-to-model flow and turbine dynamics, these surrogate models need constant calibration. In this paper, a novel real-time (joint state-parameter) estimation solution for a medium-fidelity dynamical wind farm model is presented. In this work, we demonstrate the estimation of the freestream wind speed, local turbulence, and local wind field in a two-turbine wind farm using exclusively turbine power measurements. The estimator employs an Ensemble Kalman filter with a low computational cost of approximately 1.0 s per timestep on a dual-core notebook CPU. This work presents an essential building block for real-time wind farm control using computationally efficient dynamical wind farm models.

Wake steering by active yawing of upstream wind turbines is a promising wind plant control technique. To enable the development of model-based wind plant control methods, there is a need for models that can marry the contrasting requirements of good fidelity and low computational cost. This paper presents a reduced-order model (ROM) obtained by directly compressing high-fidelity computational fluid dynamics (CFD) simulation data using the proper orthogonal decomposition (POD) method. At first, simulations of wake-interacting wind turbines are obtained for time-varying yaw settings using the lifting-line large-eddy simulation (LES) code SOWFA. Next, a ROM is synthesized from the CFD transient simulations, obtaining a discrete-time state-space model that captures the dominant dynamics of the underlying high-fidelity model with only a reduced number of states. The ROM is optionally augmented with a Kalman filter, which feeds back turbine power measurements from the plant to the model, enhancing its accuracy. Results obtained in realistic turbulent conditions show a good agreement between high-fidelity CFD solutions and the proposed POD-based ROM in terms of wake behavior and power output of waked turbines. Additionally, the ROM presents acceptable results when compared to wind tunnel experiments, including the capability of the model to partially correct for an intentionally built-in model mismatch.

Turbine wake interactions in wind farms result in decreased power extraction in downstream rows. This work investigates dynamic induction and yaw control of wind farms for increased total power extraction. Six different wind farm layouts are considered, and the relative benefits of induction control, yaw control, and combined induction–yaw control are compared. It is found that optimal control significantly increases wind-farm efficiency for virtually all cases, and that the most profitable control strategy between yaw and induction control depends on the effective farm layout as seen by the flow, and hence the mean wind direction.

This work presents the next step in realizing lidar-based closed-loop wake redirection control. Lidar-based closed-loop wake redirection aims at repositioning the wake at a desired position by yawing the wind turbine. The actual wake deflection is derived from lidar measurements and used in a closed-loop control scheme. Compared to an open-loop setting in which temporal changes are not taken into account, lidar-based closed-loop wake redirection can react on temporal disturbances. This yields a more robust control solution due to the employed closed-loop control framework. In this work, for the first time, the concept is implemented in an LES environment namely the PArallelized Large-eddy simulation Model (PALM) code. In PALM lidar measurements are simulated using a lidar model which are processed to estimate the wake position. A controller is synthesized by the usage of a the reduced order wind farm model WindFarmSimulator (WFSim). High-fidelity simulation results illustrate the controller's ability to adapt to a temporal changing crosswind disturbance in a turbulent simulation scenario. Consequently, it increases the power output of the two-turbine scenario compared to the open-loop approach.

Intelligent wind park monitoring systems may allow cutting the levelized cost of wind-generated electricity by deploying maintenance personnel more efficiently. The non-contacting passive radar technology and advanced sensing technologies on the plant side offer significant potential for such monitoring systems. The goal of the 3-year-long project ISO.Wind is to identify the most cost-efficient sensing technologies to detect maintenance-relevant damages and to use them for a wind park monitoring system. For this purpose a commercial 3MW wind turbine is instrumented with strain gauges following IEC standard 61400-13 and a network of accelerometers. It is also monitored by passive radar technology. A learning algorithm is developed and fed with available data from the sensor systems and operational data from the instrumented wind turbine. The algorithm is capable of detecting operational patterns and damage cases of the wind turbine. A graphic user interface illustrates these conditions in a comprehensible way. First field measurements show the suitability of the passive radar technology to detect the damage-relevant dynamics of the instrumented wind turbine. Validated simulations of typical damage cases prove that both instrumentation on the plant side and the passive radar sensing technology allow reliable damage detection for the examined wind turbine.

In this paper, an active power control (APC) approach for wind farms is studied, for a case in which the wind turbines interact with each other aerodynamically through their wakes. We demonstrate that the structural loadings on the individual wind turbines can be coordinated to expand their lifetimes, while the wind farm production tracks a power reference signal. We propose an additional feedback control loop in order to adjust the distribution of the regulated power demands among the wind turbines, exploiting the non-unique solutions of APC for wind farms. The axial induction factor of each wind turbine is considered as a control input to influence the overall wind farm performance. The applicability of the controller is tested with a wind farm example consisting of 2×3 turbines with partial wake overlaps and detailed interactions with the atmospheric boundary layer, simulated with the PArallelized Large-eddy simulation Model (PALM). The results demonstrate the effectiveness of the proposed approach and point to some future studies that may improve and extend the performance.

The objective of this paper is to incorporate sparse sensor data to improve flow-field estimates in a wind farm, which can then be used to perform better online wind farm optimization and control. A sparse-sensing algorithm is used to determine the optimal locations of sensors to improve the overall estimation precision of the flow field within the wind farm. This algorithm takes advantage of the dominant atmospheric structures in a wind farm to reconstruct the flow field from point measurements in the field. These measurements, in their optimal locations, have the ability to improve the observability of a wind farm and thus provide faster, more accurate, state estimation.

As the wind energy penetration increases, it is becoming more important for wind farms to contribute to frequency regulation, i.e. active power control on the power grid. However, before such controllers can be widely used in industry, they need to be thoroughly validated. This paper contributes to this task by presenting an experimental setup suitable for validation of wind farm control algorithms in wind tunnel experiments. A model free closed loop active power control is implemented on model wind turbines, and tested in a wind tunnel at University of Oldenburg under different operating conditions. Obtained results show that the proposed controller is sufficient for following power references, which is in accordance to numerical studies available in the literature. It is also verified that the controller can react to different disturbances, including variations in the wind farm power reference, and shut-downs of certain turbines in the wind farm. Furthermore, the possibility of coupling the proposed controller with control algorithms for reduction or fairer distribution of structural toads in the wind farm is also demonstrated.

Modern wind turbines have access to highly reliable measurements of important control input signals, such as generator speed and nacelle acceleration. They also have high fidelity numerical models such as in Bladed, which can be used to estimate structural loads under simulated normal and extreme operating conditions. However, if we want to know the structural loads that occurred in a time period on the real turbine, presently this requires instrumentation with strain gauges. These sensors can be unreliable and expensive to install, calibrate and maintain. The price of reliable sensors is unlikely to drop to an affordable level for onshore wind in the near future. This paper describes a method of fusing the control input signals with the turbine numerical model to estimate structural loads online in real time. The estimator is validated in Bladed simulations of a Goldwind 6 MW turbine.

We discuss the detection of rotational periodic torque deviations in variable speed wind turbine systems. These deviations can be caused by faults in the system. The turbine torque is estimated with an observer and an estimate of the ideal aerodynamical torque is calculated. These torques are analysed using a phase-locked loop to detect deviations. The torque observer is based on the model of the turbine drive train. It is modelled as a two-mass-system with a flexible shaft. The design of the observer and the phase-locked loop are shown and their stability is discussed. Simulations show, that the presented concept is capable of detecting the amplitude of the deviations at different periodicities, online and for variable speed.

The objective of active power control in wind farms is to provide ancillary grid services. Improving this is vital for a smooth wind energy penetration in the energy market. One of these services is to track a power reference signal with a wind farm by dynamically de- and uprating the turbines. In this paper we present a computationally efficient model predictive controller (MPC) for computing optimal control signals for each time step. It is applied in the PArallelized Large-eddy simulation Model (PALM), which is considered as the real wind farm in this paper. By taking measurements from the PALM, we show that the closed-loop controller can provide power reference tracking while minimizing variations in the axial forces by solving a constrained optimization problem at each time step. A six turbine simulation case study is presented in which the controller operates with optimised turbine yaw settings. We show that with these optimized yaw settings, it is possible to track a power signal that temporarily exceeds the power harvested when operating under averaged greedy control turbine settings. Additionally, variations in the turbine's force signals are studied under different controller settings.

In this paper it is shown that measured data in a wind turbine, available to the controller, can be formulated into a polynomial regression problem in order to estimate the turbine's maximum efficiency power coefficient, C_p,max, and drivetrain losses, assuming the latter can be well approximated as being linear. Gaussian process (GP) machine learning is used for the regression problem. These formulations are tested on data generated using the Supergen Exemplar 5 MW wind turbine model, with results indicating that this is a potential low cost method for detecting changes in aerodynamic efficiency and drivetrain losses. The GP approach is benchmarked against standard least-squares (LS) regression, with the GP shown to be the superior method in this case.

Lidar-assisted control of wind turbines has been an active field of research during the last years and has recently become more attention from industry, as well. The potential of lidar-assisted feed-forward controllers is shown in various simulation studies and proven in first field-testings. This work aims to further push forward the application of lidar-assisted control by introducing a method to bring the wind turbine to a different operating point, when a lidar detects an approaching extreme gust. The method reduces the power output of the wind turbine while keeping the rotor rotation constant. This is achieved by a synchronous multi-variable feed-forwarding of the generator torque and the pitch angle. In combination with a classical lidar-assisted feed-forward controller, this leads to further reduced structural loads under the impact of an extreme gust. The method is tested with a coherent and a corresponding turbulent gust.

Control and operation systems of wind turbines must primarily ensure the fully automatic operation of wind turbines in a constantly changing environment. Economic efficiency charges the control system to ensure that the highest possible efficiency is achieved and the mechanical loads caused by disturbances are minimized. The ability of an observer, in this case a Kalman filter (Kf), to estimate non-measurable states from a set of measurements using a model of the plant suggests the idea of extending the model of the plant by a model of the disturbance. Disturbance states thus can be reconstructed and an easy-to-determine quasi-disturbance-feedforward controller can be used to reject them. This method is called Disturbance-Accommodating Control (DAC). In this paper, Dryden's turbulence model-which shapes a white noise signal via a form filter to meet spectrum conditions - and an inverse notch filter to model the rotational sampling effect are used for each blade, in contrary to the hitherto used deterministic disturbance models or the simple random walk models for stochastic turbulence. Measurement- and model-uncertainties are described as uncorrelated white noise. With this approach, the requirements of the Kf derivation are met and quantitative measures for the Kf process noise covariance matrix are available especially for the disturbance. The simplified tuning process and the high potential for load reduction are demonstrated for the NREL 5 MW Wind turbine. The reduction by a factor of 4.4 of the standard deviation of the flapwise root bending moment shows the high potential of this stochastic DAC approach. A parameter study to determine the influence of the turbulence spectrum bandwidth and to identify the dependency of the stochastic DAC approach on uncertainties of the process noise covariance matrix was performed. The study shows that the Kf is robust against a wide spectrum of parameter variations. Only if the time constant of the Dryden filter is significantly reduced, the performance is decreased.

In this paper a model based approach for condition monitoring and diagnosis of faults in the wind turbine rotor blades is described. The investigation is focused on creating a fault-free reference system, the design and optimization of the controller/observer based on the Disturbance Accommodating Control theory and the multi-objective genetic optimization algorithm in order to further enhance the condition monitoring and faults diagnosis. Test cases are presented by injecting a parameter change, e.g. pitch angle, as an error to the non-linear wind turbine simulation model in order to demonstrate that the proposed method is able to detect the fault on the blades.

Permanent magnet synchronous machines have gained popularity in wind turbines due to their merits of high efficiency, power density, and reliability. The wind turbines normally work in a wide range of operations, and harsh environments, so unexpected faults may occur and result in productivity losses. The common faults in the permanent magnet machines occur in the bearing and stator winding, being mainly detected in steady-state operating conditions under constant loads and speeds. However, variable loads and speeds are typical operations in wind turbines and powertrain applications. Therefore, it is important to detect bearing and stator winding faults in variable speed and load conditions. This paper proposes an algorithm to diagnose multiple faults in variable speed and load conditions. The algorithm is based on tracking the frequency orders associated with faults from the normalised order spectrum. The normalised order spectrum is generated by resampling the measured vibration signal via estimated motor speeds. The fault features are then generated from the tracking orders in addition to the estimated torque and speed features. Finally, support vector machine algorithm is used to classify the faults. The proposed method is validated using experimental data, and the validated results confirm its usefulness for practical applications.

Permanent magnet synchronous motors become popular in wind turbines and industrial applications. In critical machines, it is necessary to use robust condition monitoring and fault diagnosis algorithms to prevent faults or shutdowns. The data-driven approach with machine learning algorithms is widely used in industrial and research communities as this method does not require a mathematical model of the system, which is difficult to obtain in practical cases. Most of the successful machine learning methods are based on supervised learning approach, requiring labelled training data. The supervised learning approach cannot use the unlabelled data, while only a few labelled data is in place in the industry. This work uses a deep autoencoder based unsupervised learning method to identify the features of the fault classification algorithm in a self-supervised way, which overcome the shortage of labelled data. The proposed algorithm uses the benefits of available unlabelled data, but it needs only a few labelled data. The fault classification algorithm is based on artificial neural network SoftMax layer and Bayes classifier. The robustness of the algorithm is improved by fusing the current and vibration information. Experimental results are used to validate the robustness of proposed algorithms under noise conditions, and the results show that the algorithm could classify faults robustly.

Increasing demands in decentralized power plants have focused attention on Vertical Axis Wind Turbines (VAWTs). However, accessing high range of power from VAWTs is an impediment due to increased loads on the turbine blades. Here, we derive an optimal pitching action that reduces the periodic disturbance on turbine blades of VAWTs without affecting their power production. A control technique called Subspace Predictive Repetitive Control (SPRC) alongwith a LQ Tracker is used for recursive identification to estimate the parameters of VAWT model and further provide an optimal control law accordingly. Basis functions have been used to reduce the dimensionality of the control problem. Simulation results show a great potential of the data-driven SPRC approach coupled with LQ Tracker in reducing the turbine loads on VAWTs.

A novel strategy to reduce unwanted swings and motions in floating wind turbines is presented. At above rated wind speeds, the platform, on which the wind turbine is mounted, causes the generator speed control loop to become unstable. The proposed strategy assures stability of the control loop by an additive adjustment of the measured generator speed using tower fictitious forces. The developed strategy is independent of the platform and wave dynamics.

In the present contribution, the modelling of the aerodynamic coefficients of wind turbines are obtained by using an artificial neural network (ANN) and it is analysed from the control application point of view. The obtained results show that the artificial neural network approach is appropriate to fit the data of the aerodynamic coefficients with high accuracy. The advantage of the approach is that the ANN provides an analytical equation (including its derivatives) that can be embedded in the general dynamic model, which is used for the control system design.

Based on actual damage to shaft couplings and elastic supporting caused by drive train vibration which occurred in a low wind speed mountainous wind farm in China, this paper proposes a novel drive train damping method. Instead of a common lead compensator, a delay compensator is used for phase correction. Features of both methods are analysed and compared. Compared to common dampers, the proposed one is only effective at centre frequency of damping, and will not magnify responses at other frequencies, especially high frequencies. Simulations with different configurations were implemented to verify effects of the proposed method. A field test has been implemented on Shanghai Electric 2MW wind turbine. The test result shows that the proposed method is effective to solve drive train vibration, and can be extended to large-rotor turbines located in slow-speed and mountainous areas in China.

Nowadays, wind power plants (WPPs) should be able to dynamically change their power output to meet the power demanded by the transmission system operators. When the wind power generation exceeds the power demand, the WPP works in de-loading operation keeping some power reserve to be delivered into the grid to balance the frequency drop. This paper proposes to cast a model predictive control strategy as a multi-objective optimization problem which regulates the power set-points among the turbines in order to track the power demand profile, to maximize the power reserve, as well as to minimize the power losses in the inter-arrays connecting the wind turbines within the wind farm collection grid. The performance of the proposed control approach was evaluated for a wind farm of 12 turbines using a wind farm simulator to model the dynamic behavior of the wake propagation through the wind farm.

While substructures of offshore wind turbines become older and begin to reach their design lifetimes, the relevance of measurement based lifetime extension increases. To make well-founded decisions on possible lifetime extensions, damage extrapolations based on measurements are needed. However, although for all substructures, fatigue damage calculations were conducted during the design process, there is no consensus on how to extrapolate 10-minute damages to lifetime damages. Furthermore, extrapolating damages is an uncertain process and its actual reliability is unknown. Therefore, the current work uses data of offshore strain measurements to assess different approaches of extrapolating damages, and to investigate the reliability of damage extrapolations. For the present data, the most reliable lifetime estimations are possible, if the damage data is split up into wind speed bins. For each wind speed bin, the occurrence probability should be based on data rather than on design documents. Moreover, using mean damages in each bin is the best practice. Furthermore, our results suggest that strain measurements of about 9 to 10 months lead to a relatively representative and unbiased data set. Therefore, if there are no significant changes of the turbine or the environmental conditions over the lifetime, damage extrapolations based on such a time period are sufficiently accurate.

Failures in wind turbine pitch systems can cause significant outages in offshore wind turbines due to the finite weather windows for maintenance. In a complex system such as the pitch system, a fault in the gearbox can contaminate the motor current signals and result in a misdiagnosis. This paper investigates a sensor fusion technique to reliably diagnose faults in pitch motors and multistage planetary gearboxes in pitch drives. A support vector machine classifier is used for fault diagnosis based on features extracted from motor currents and gearbox vibration signals. The approach is validated with three commonly occurring pitch drive faults, namely, the stator turns fault in the pitch motor, input shaft bearing and planet gear fault in the planetary gearbox. The developed diagnostic method is validated with artificially seeded faults in a laboratory setup of a scaled pitch drive.

In recent years, Light Detection and Ranging (LiDAR) has emerged as a feasible, reliable and accurate remote sensing technology for wind speed measurements. For a fine and performant use of the LiDAR data in turbine control and monitoring applications, adapted and sophisticated real-time processing is needed. With the objective to retrieve accurate enhanced information from the sensor raw data, this paper aims to present solution to estimate, in three dimensions and in real time, incoming wind characteristics, such as wind speed, wind direction and instantaneous shears. An innovative reconstruction is proposed based on recursive weighted least squares method. The approach is validated for pulsed nacelle LiDAR systems with simulated data, obtained with a specifically developed wind generator tool, including a representative LiDAR Wind Iris TC model.

Operational data from wind farms is crucial for wind turbine condition monitoring and performance assessment. In this paper, we analyse three wind farms with the aim to monitor environmental and operational conditions that might result in underperformance or failures. The assessment includes a simple wind speed characterisation and wake analysis. The evolution of statistical parameters is used to identify anomalous turbine behaviour. In total, 88 turbines and 12 failures are analysed, covering different component failures. Notwithstanding the short period of data available, several operational parameters are found to deviate from the farm trend in some turbines affected by failures. As a result, some parameters show better monitoring capabilities than others, for the detection of certain failures. However, the limitations of SCADA statistics are also shown as not all failures showed anomalies in the observed parameters.

Derating of individual turbines is one of the options to implement wind farm performance optimization, although there are different ways to proceed with such derating at the turbine control level. The present paper develops an option based on the minimal thrust coefficient in accordance with the C_p and C_t contour levels. This strategy is compared for the region below rated wind speed with two other strategies where either the pitch or the tip speed ratio are maintained at their maximum C_p values from normal operation. The study concludes that maintaining the pitch at the optimal value from normal operation produces poorer performance from the thrust and the loads perspective. Practical implementation issues have also been detected.

The use of down-regulation or curtailment control strategies for wind turbines offers means of supporting the stability of the power grid and also improving the efficiency of a wind farm. Typically, wind turbine derating is performed by modifying the power set-point and subsequently, the turbine control input, namely generator torque and blade pitch, are acted on to such changes in the power reference. Nonetheless, in addition to changes in the power reference, derating can be also performed by modifying the rotor speed set-point. Thus, in this work, we investigate the performance of derating strategies with different rotor speed set-point, and in particular, their effect on the turbine structural fatigue and thrust coefficient were evaluated. The numerical results obtained from the high-fidelity turbine simulations showed that compared to the typical derating strategy, the derated turbines might perform better with lower rotor speed set-point but it is crucial to ensure such a set-point does not drive the turbine into stalled operations.

A typical concern in rotating systems is related to rotor imbalances, which result typically from pitch misalignment and unbalanced mass distribution. A novel control for simultaneously targeting mass and pitch imbalances on the rotor is presented. Additionally, a novel detection strategy is developed in order to detect the imbalance source out of the behavior of the control action. More in depth, since the control will generate an artificial aerodynamic imbalance which compensates the pre-existent aerodynamic and inertial ones, one can find and interpret the fingerprint of the imbalance source in the behavior of the balancing controller. A clear advantage of this approach is that the imbalance detection is performed while the control keeps the machine working within its operating limits reducing the down-time and unscheduled maintenance actions.

In the paper, the potential to alleviate wind turbine loads through combined implementation of different passive and active control methods is assessed. Passive control of loads is accomplished through blade designs with build in material and/or geometric bend-twist coupling (BTC). The first is materialized by introducing an offset angle on the plies of the uni-directional material over the spar caps of the blade, while the latter by sweeping the blade elastic axis with respect to the pitch axis. Active control of loads is considered through individual pitch control (IPC) or concurrent use of individual pitch and flap control (IPC+IFC). Different combinations of the abovementioned techniques are tested in the paper with the aim to obtain maximum possible load reduction levels but also confine key design parameters of the various methods within reasonable limits that by no means exceed manufacturing constraints. The performance of the different control options is assessed through aeroelastic simulations for the 10MW DTU Reference Wind Turbine (RWT). A subset of representative fatigue and ultimate design load cases (DLCs) of the IEC is simulated and load reduction levels are assessed with respect to the baseline RWT configuration with no aeroelastic control of loads.

This paper proposes a robust fault detection and isolation (FDI) technique for the power electronic converter (PEC) of doubly-fed induction generator (DFIG) wind turbines (WTs), and in particular for open-circuit faults herein. It combines fault indicators based on the processing of the Clarke transformation of the converter currents and a statistical change detection algorithm, namely a cumulative sum (CUSUM) algorithm that detects significant changes in the variance of the reactive power. This allows for a reduction of the false alarm rate compared to an approach relying exclusively on the current analysis. The proposed FDI technique is validated by means of both simulation and experimental results.

Effectively monitoring the health of a wind turbine gearbox is a complex and often multidisciplinary endeavor. Recently, condition monitoring practices increasingly combine knowledge from fields like signal processing, machine learning, and mechanics. Such a diverse approach becomes necessary when dealing with the vast amount of data that is generated by the multitude of sensors that are typically placed on a wind turbine gearbox. Ideally, this approach needs to be automated and scalable as well, since it is unfeasible to perform all the necessary processing work manually in a continuous manner. This paper focuses on assessing the performance of such an automated processing framework for the case of gearbox fault detection using vibration measurements. A year of vibration measurements on a gearbox is simulated by stochastic variation of the operating conditions and the system behavior. A bearing fault is progressively introduced as to track the detection capabilities of the framework in such stochastic circumstances. The used signal model is based on previously obtained experience with experimental data sets originating from wind turbine gearboxes. The framework itself consists of multiple pre-processing steps where each step tries to deal with compensating for the external or unwanted influences such as speed variation or noise. Finally, multiple features are calculated on the pre-processed signals and trended as to see whether the processing scheme can provide any benefit compared to basic traditional statistical indicators. It is shown that the multi-step pre-processing approach is beneficial and robust for the advanced feature calculation and thus the early fault detection.

In the paper, an individual flap control (IFC) algorithm based on wind speed measurements obtained with a spinner anemometer is presented. The controller uses flap actuators with the aim to remove any deterministic source of load variation on blades, associated with asymmetries of the inflow. Such load variations are concentrated on multiples of the rotational frequency (p multiples) and they are mainly due to wind yaw, inclination and atmospheric boundary layer (ABL) shear. The aim of the controller is to assist operation of the conventional feed-back individual pitch controller (IPC) and thereby reduce its control duty cycle. The performance of the proposed controller is assessed through aero-elastic simulations for the 10MW DTU Reference Wind Turbine. A subset of normal operation fatigue and ultimate Design Load Cases (DLCs) of the IEC has been simulated and load reduction capabilities of the control algorithm have been compared against those of the standard individual pitch control (IPC) loop.

The upstream wind is known as the main source of disturbance to the wind turbine. Therefore, having the wind information before it hits the turbine, allows the wind turbine controller to take necessary actions to proper rejection of the disturbances. Several advanced control methods have been proposed to exploit the LIDAR wind measurements to enhance the wind turbine control. To date, the Nonlinear Model Predictive Control (NMPC), has been one of the most successful methods in using wind data to improve the control performance. However, due to the immense computational burden, its real-time application is still challenging. Very recently, an implementation of the Exact Output Regulation (EOR) scheme for wind turbines control has been proposed. In this paper, the performances of the two model-based controllers, namely the NMPC controller and the EOR controller, in mitigating mechanical loads on the DTU10MW wind turbine are compared against each other. The results are also compared against the classic baseline feedback controller. Simulation results indicate that both controllers show significant improvement in reducing fatigue loads on the wind turbine structure, whilst maintaining the power production at the desired rated level in comparison to the baseline controller. It is also shown that the EOR controller can closely compete with the disturbance rejection performance of the NMPC controller. The simulation running times are considerably lower for the EOR scheme, potentially making EOR more suitable for real-time applications.