Preview predictive control layer design based upon known wind turbine blade‐pitch controllers

The use of upstream wind measurements has motivated the development of blade‐pitch preview controllers for improving rotor speed tracking and structural load reduction beyond that achievable via conventional feedback control. Such preview controllers, typically based upon model predictive control (MPC) for its constraint handling properties, alter the closed‐loop dynamics of the existing blade‐pitch feedback control system. This can result in a deterioration of the robustness properties and performance of the existing feedback control system. Furthermore, performance gains from utilising the upcoming real‐time measurements cannot be easily distinguished from the feedback control, making it difficult to formulate a clear business case for the use of preview control. Therefore, the aim of this work is to formulate a modular MPC layer on top of a given output‐feedback blade‐pitch controller, with a view to retaining the closed‐loop robustness and frequency‐domain performance of the latter. The separate nature of the proposed controller structure enables clear and transparent quantification of the benefits gained by using preview control, beyond that of the underlying feedback controller. This is illustrated by results obtained from high‐fidelity closed‐loop turbine simulations, showing the proposed control scheme incorporating knowledge of the oncoming wind and constraints achieved significant 43% and 30% reductions in the rotor speed and flap‐wise blade moment standard deviations, respectively. Additionally, the chance of constraint violations on the rotor speed decreased remarkably from 2.15% to 0.01%, compared to the nominal controller. Copyright © 2017 John Wiley & Sons, Ltd.     

Sensor comparison study for load alleviating wind turbine pitch control

As the size of wind turbines increases, the load alleviating capabilities of the turbine controller are becoming increasingly important. Load alleviating control schemes have traditionally been based on feedback from load sensor; however, recent developments of measurement technologies have enabled control on the basis of preview measurements of the inflow acquired using, e.g., light detection and ranging. The potential of alleviating load variations that are caused by mean wind speed changes through feed‐forward control have been demonstrated through both experiments and simulations in several studies, whereas the potential of preview control for alleviating the load variations caused by azimuth dependent inflow variations is less described. Individual or cyclic pitch is required to alleviate azimuth dependent load variations and is traditionally applied through feedback control of the blade root loads. In many existing studies, the performance of an advanced controller is compared with the performance of a simpler controller. In this study, the effect of three measurement types on the load alleviating performance of the same cyclic pitch control design is studied. By using a baseline cyclic pitch controller as test bench, the effect of the different measurement types on the controller performance can be assessed independent of control design. The three measurement types that are considered in this study are as follows: blade root out‐of‐plane bending moment, on‐blade measurements of angle of attack and relative velocity at a radial position of the blades, and upstream inflow measurements from a spinner mounted light detection and ranging (LiDAR) sensor that enables preview of the incoming flow field. The results show that for stationary inflow conditions, the three different measurement types yield similar load reductions, but for varying inflow conditions, the LiDAR sensor‐based controller yields larger load reductions than the two others. The results also show that the performance of the LiDAR sensor‐based controller is very sensitive to uncertainties relating to the inflow estimation. Copyright © 2013 John Wiley & Sons, Ltd.     

风电机组模糊PI控制与前馈控制相结合的变桨控制

针对经典PID变桨距控制器超调量大、波动剧烈等缺陷,提出了一种采用模糊PI控制与前馈控制相结合的控制器来控制浆距角,进而使机组功率稳定在额定功率附近的控制思路。通过对某1.5MW风力发电机组进行仿真,结果表明,该控制器控制效果优于经典PID变桨距控制器。     

Comparison and testing of power reserve control strategies for grid‐connected wind turbines

The stability of the electrical grid depends on enough generators being able to provide appropriate responses to sudden losses in generation capacity, increases in power demand or similar events. Within the United States, wind turbines largely do not provide such generation support, which has been acceptable because the penetration of wind energy into the grid has been relatively low. However, frequency support capabilities may need to be built into future generations of wind turbines to enable high penetration levels over approximately 20%. In this paper, we describe control strategies that can enable power reserve by leaving some wind energy uncaptured. Our focus is on the control strategies used by an operating turbine, where the turbine is asked to track a power reference signal supplied by the wind farm operator. We compare the strategies in terms of their control performance as well as their effects on the turbine itself, such as the possibility for increased loads on turbine components. It is assumed that the wind farm operator has access to the necessary grid information to generate the power reference provided to the turbine, and we do not simulate the electrical interaction between the turbine and the utility grid. Copyright © 2013 John Wiley & Sons, Ltd.     

Nacelle anemometer measurement‐based extremum‐seeking wind turbine region‐2 control for improved convergence in fluctuating wind

For region‐2 operation of wind turbines in practice, the optimal torque gain can deviate from the nominal value because of the variations in turbine and wind conditions. The extremum‐seeking control (ESC) has shown its potential as a model‐free region‐2 control solution in some recent work; however, the ESC with rotor power feedback suffers from undesirable convergence under fluctuating wind. In this paper, we propose to use an estimated power coefficient as the objective function for the torque‐gain ESC, where the hub‐height free‐stream wind speed (FSWS) is estimated with the nacelle anemometer measurement on the basis of the so‐called nacelle transfer function (NTF) between the nacelle anemometer and met‐tower measurement. A sensitivity analysis is performed to quantify the impact of the wind speed estimation error on the estimation of power coefficient. An ESC integrated interregion switching scheme is proposed to avoid the load increase. Simulation results show that, compared with the power feedback‐based ESC, the proposed method can greatly improve the convergence rate of ESC under fluctuating wind, even under relatively large wind speed estimation error. Evaluation for the fatigue loads of wind turbine shows that the proposed control strategy induces mild increase of the wind turbine load.     

Response analysis and comparison of a spar‐type floating offshore wind turbine and an onshore wind turbine under blade pitch controller faults

This paper analyses the effects of three pitch system faults on two classes of wind turbines, one is an onshore type and the other a floating offshore spar‐type wind turbine. A stuck blade pitch actuator, a fixed value fault and a bias fault in the blade pitch sensor are considered. The effects of these faults are investigated using short‐term extreme response analysis with the HAWC2 simulation tool. The main objectives of the paper are to investigate how the different faults affect the performance of wind turbines and which differences exist in the structural responses between onshore and floating offshore wind turbines. Several load cases are covered in a statistical analysis to show the effects of faults at different wind speeds and fault amplitudes. The severity of individual faults is categorized by the extreme values the faults have on structural loads. A pitch sensor stuck is determined as being the most severe case. Comparison between the effects on floating offshore and onshore wind turbines show that in the onshore case the tower, the yaw bearing and the shaft are subjected to the highest risk, whereas in the offshore case, the shaft is in this position. Copyright © 2014 John Wiley & Sons, Ltd.     

A flywheel in a wind turbine rotor for inertia control

In this paper, a flywheel energy storage that is an integral part of a wind turbine rotor is proposed. The rotor blades of a wind turbine are equipped with internal weights, which increase the inertia of the rotor. The inertia of this flywheel can be controlled by varying the position of the weights, i.e. by positioning them closer to the center of rotation (closer to the hub) or closer to the tip of the blades. The simulation model used in this study is introduced briefly. The equation system of the flywheel is set up. Finally, simulations of different scenarios show the performance of this controllable flywheel. The conclusion is that the proposed system can mitigate transients in the power output of wind turbines. Hence, it can support the frequency control in a power system by contributing to the power system inertia. © 2014 The Authors. Wind Energy published by John Wiley & Sons, Ltd.     

Plasma‐based feed‐forward dynamic stall control on a vertical axis wind turbine

Dynamic stall was controlled on a double‐bladed H‐Rotor vertical axis wind turbine model using pulsed dielectric barrier discharge plasma actuators in a feed‐forward control configuration. The azimuthal angles of plasma actuation initiation and termination, that produced the largest increases in power, were determined parametrically on the upstream half of the turbine azimuth in a low‐speed blow‐down wind tunnel at wind speeds of 7 m/s. A mathematical model, together with instantaneous turbine speed, was used to estimate transient torque and power developed by the turbine under the influence of plasma actuation. Overall performance improvements were based on changes between the final actuated and initial baseline results. A remarkable result of this investigation was that a net turbine power increase of 10% was measured. This was achieved by systematically reducing plasma pulsation duty cycles as well as the plasma initiation and termination angles. Nevertheless, it was determined that further performance increases could be achieved by changing the actuator's dielectric material, increasing the turbine radius and developing a method for control of dynamic stall on both the upwind (inboard of the blades) and downwind (outboard of the blades) halves of the turbine azimuth. Copyright © 2014 John Wiley & Sons, Ltd.     

Power generation control of a hydrostatic wind turbine implemented by model‐free adaptive control scheme

The hydrostatic wind turbine (HWT) is a type of wind turbine that uses hydrostatic transmission (HST) drivetrain to replace the traditional gearbox drivetrain. Without the fragile and expensive gearbox and power converters, HWT can potentially reduce the maintenance costs owing to the gearbox and power converter failures in wind power system, especially in offshore cases. We design an MFAC torque controller to regulate the pump torque of the HWT and compared with an torque controller. Then we design an MFAC pitch controller to stabilise the rotor speed of HWT and compared with a gain‐scheduling proportional‐integral (PI) controller and a gain‐scheduling PI controller with antiwindup (PIAW). The results indicate that MFAC torque controller provides more effective tracking performance than the controller and that MFAC pitch controller shows better rotor speed stabilisation performance in comparison with the gain‐scheduling PI controller and PIAW.     

Model‐based control addition to prescribe DFIG wind turbine fast frequency response

This paper investigates the physical capability of double‐fed induction generator (DFIG) wind turbines for inertial support of frequency response. Frequency stability is modeled using the DFIG electromechanical and generator controller dynamics, and a destabilizing effect is demonstrated in low‐inertia systems. To improve response, a synchronous reference frame DFIG controller is proposed that acts by following low‐frequency grid dynamics and adds a fast acting proportional plus integral (PI)‐controlled frequency‐responsive component to existing qd current commands. The proposed controller is derived in a straightforward manner using only the DFIG dynamic equations and is designed using pole/zero placement techniques. Laboratory experiments using a micro‐scale DFIG wind turbine with hub‐emulating flywheel prove better capability for transient frequency regulation even under extreme load change. The result is a DFIG controller that balances the appearance of transients in electrical and mechanical systems. Value is achieved in providing immediate continuous inertial response to support load change. The proposed frequency response can improve the use of existing physical inertia from wind turbines.     

Copula‐based model for wind turbine power curve outlier rejection

Power curve measurements provide a conventional and effective means of assessing the performance of a wind turbine, both commercially and technically. Increasingly high wind penetration in power systems and offshore accessibility issues make it even more important to monitor the condition and performance of wind turbines based on timely and accurate wind speed and power measurements. Power curve data from Supervisory Control and Data Acquisition (SCADA) system records, however, often contain significant measurement deviations, which are commonly produced as a consequence of wind turbine operational transitions rather than stemming from physical degradation of the plant. Using such raw data for wind turbine condition monitoring purposes is thus likely to lead to high false alarm rates, which would make the actual fault detection unreliable and would potentially add unnecessarily to the costs of maintenance. To this end, this paper proposes a probabilistic method for excluding outliers, developed around a copula‐based joint probability model. This approach has the capability of capturing the complex non‐linear multivariate relationship between parameters, based on their univariate marginal distributions; through the use of a copula, data points that deviate significantly from the consolidated power curve can then be removed depending on this derived joint probability distribution. After filtering the data in this manner, it is shown how the resulting power curves are better defined and less subject to uncertainty, whilst broadly retaining the dominant statistical characteristics. These improved power curves make subsequent condition monitoring more effective in the reliable detection of faults. Copyright © 2013 John Wiley & Sons, Ltd.     

A combined aeroelastic‐aeroacoustic model for wind turbine noise: verification and analysis of field measurements

In this paper, semi‐empirical engineering models for the three main wind turbine aerodynamic noise sources, namely, turbulent inflow, trailing edge and stall noise, are introduced. They are implemented into the in‐house aeroelastic code HAWC2 commonly used for wind turbine load calculations and design. The results of the combined aeroelastic and aeroacoustic model are compared with field noise measurements of a 500 kW wind turbine. Model and experimental data are in fairly good agreement in terms of noise levels and directivity. The combined model allows separating the various noise sources and highlights a number of mechanisms that are difficult to differentiate when only the overall noise from a wind turbine is measured. Copyright © 2017 John Wiley & Sons, Ltd.     

Using nacelle‐based wind speed observations to improve power curve modeling for wind power forecasting

Wind power forecasting for projection times of 0–48 h can have a particular value in facilitating the integration of wind power into power systems. Accurate observations of the wind speed received by wind turbines are important inputs for some of the most useful methods for making such forecasts. In particular, they are used to derive power curves relating wind speeds to wind power production. By using power curve modeling, this paper compares two types of wind speed observations typically available at wind farms: the wind speed and wind direction measurements at the nacelles of the wind turbines and those at one or more on‐site meteorological masts (met masts). For the three Australian wind farms studied in this project, the results favor the nacelle‐based observations despite the inherent interference from the nacelle and the blades and despite calibration corrections to the met mast observations. This trend was found to be stronger for wind farm sites with more complex terrain. In addition, a numerical weather prediction (NWP) system was used to show that, for the wind farms studied, smaller single time‐series forecast errors can be achieved with the average wind speed from the nacelle‐based observations. This suggests that the nacelle‐average observations are more representative of the wind behavior predicted by an NWP system than the met mast observations. Also, when using an NWP system to predict wind farm power production, it suggests the use of a wind farm power curve based on nacelle‐average observations instead of met mast observations. Further, it suggests that historical and real‐time nacelle‐average observations should be calculated for large wind farms and used in wind power forecasting. Copyright © 2011 John Wiley & Sons, Ltd.     

Performance enhancement of the artificial neural network–based reinforcement learning for wind turbine yaw control

The yaw angle control of a wind turbine allows maximization of the power absorbed from the wind and, thus, the increment of the system efficiency. Conventionally, classical control algorithms have been used for the yaw angle control of wind turbines. Nevertheless, in recent years, advanced control strategies have been designed and implemented for this purpose. These advanced control strategies are considered to offer improved features in comparison to classical algorithms. In this paper, an advanced yaw control strategy based on reinforcement learning (RL) is designed and verified in simulation environment. The proposed RL algorithm considers multivariable states and actions, as well as the mechanical loads due to the yaw rotation of the wind turbine nacelle and rotor. Furthermore, a particle swarm optimization (PSO) and Pareto optimal front (PoF)‐based algorithm have been developed in order to find the optimal actions that satisfy the compromise between the power gain and the mechanical loads due to the yaw rotation. Maximizing the power generation and minimizing the mechanical loads in the yaw bearings in an automatic way are the objectives of the proposed RL algorithm. The data of the matrices Q (s,a) of the RL algorithm are stored as continuous functions in an artificial neural network (ANN) avoiding any quantification problem. The NREL 5‐MW reference wind turbine has been considered for the analysis, and real wind data from Salt Lake, Utah, have been used for the validation of the designed yaw control strategy via simulations with the aeroelastic code FAST.     

A kernel plus method for quantifying wind turbine performance upgrades

Power curves are commonly estimated using the binning method recommended by the International Electrotechnical Commission, which primarily incorporates wind speed information. When such power curves are used to quantify a turbine's upgrade, the results may not be accurate because many other environmental factors in addition to wind speed, such as temperature, air pressure, turbulence intensity, wind shear and humidity, all potentially affect the turbine's power output. Wind industry practitioners are aware of the need to filter out effects from environmental conditions. Toward that objective, we developed a kernel plus method that allows incorporation of multivariate environmental factors in a power curve model, thereby controlling the effects from environmental factors while comparing power outputs. We demonstrate that the kernel plus method can serve as a useful tool for quantifying a turbine's upgrade because it is sensitive to small and moderate changes caused by certain turbine upgrades. Although we demonstrate the utility of the kernel plus method in this specific application, the resulting method is a general, multivariate model that can connect other physical factors, as long as their measurements are available, with a turbine's power output, which may allow us to explore new physical properties associated with wind turbine performance. Copyright © 2014 John Wiley & Sons, Ltd.     

A compact,closed‐form solution for the optimum,ideal wind turbine

The classical momentum solution for the optimum induced‐flow distribution of a wind turbine in the presence of wake swirl can be found in many textbooks. This standard derivation consists of two momentum balances (one for axial momentum and one for angular momentum), which are combined into a formula for power coefficient in terms of induction factors. Numerical procedures then give the proper induction factors for the optimum inflow distribution at any radial station; and this, in turn, gives the best possible power coefficient for an ideal wind turbine. The present development offers a more straightforward derivation of the optimum turbine. The final formulas give the identical conditions for the ideal wind turbine as do the classical solutions—but with several important differences in the derivation and in the form of the results. First, only one momentum balance is required (the other being redundant). Second, the solution is provided in a compact, closed form for both the induction factors and the minimum power—rather than in terms of a numerical process. Third, the solution eliminates the singularities that are present in current published solutions. Fourth, this new approach also makes possible a closed‐form solution for the optimum chord distribution in the presence of wake rotation. Copyright © 2013 John Wiley & Sons, Ltd.     

Radar‐based structural health monitoring of wind turbine blades: The case of damage localization

This short communication reports on a radar approach for structural health monitoring of wind turbine blades. Therefore, a bistatic frequency‐modulated continuous wave (FMCW) radar in the frequency range from 33.4 to 36.0 GHz has been developed and tested experimentally using a laboratory wind turbine demonstrator. A differential damage localization framework is presented here that exploits signal differences between measurements from the intact and the damaged structure for 3D imaging of the defect. We have achieved the localization of a 30‐mm cut in a glass fiber composite structure as well as the localization of a water pack at the backside of the specimen with a localization error of several centimeters.     

Large‐eddy simulations with extremum‐seeking control for individual wind turbine power optimization

Large‐eddy simulations of the flow past an array of three aligned turbines have been performed. The study is focused on below rated (Region 2) wind speeds. The turbines are controlled through the generator torque gain, as usually done in Region 2. Two operating strategies are considered: (i) preset individual optimum torque gain based on a model for the power coefficient (baseline case) and (ii) real‐time optimization of torque gain for maximizing each individual turbine power capture during operation. The real‐time optimization is carried out through a model‐free approach, namely, extremum‐seeking control. It is shown that ESC is capable of increasing the power production of the array by 6.5% relative to the baseline case. The extremum‐seeking control reduces the torque gain of the downstream turbines, thus increasing the angular speed of the blades. This results in improved aerodynamics near the tip of the blade that is the portion contributing mostly to the torque and power. In addition, an increase in angular speed leads to a larger entrainment in the wake, which also contributes to provide additional available power downstream. It is also shown that the tip speed ratio may not be a reliable performance indicator when the turbines are in waked conditions. This may be a concern when using optimal parameter settings, determined from isolated turbine models, in applications with waked turbines. Copyright © 2017 John Wiley & Sons, Ltd.     

Design and laboratory test of black‐start control mode for wind turbines

In this article, an operational strategy and control concept for wind turbines (WTs) are described, which would allow them to actively contribute to black‐start events after major power system outages. The approach is based on (a) a new generator/converter control strategy implementing a so called virtual synchronous machine (VSM) and (b) a number of modifications to the superimposed WT controller allowing for operation in black‐start conditions. In order to operate stably even at very low active power levels and to cope with sudden changes in active power, due to switching of loads in the recovered grid, the rotor speed/pitch controller had to be redesigned. The extension of the operational range of the WT towards negative power, ie, power consumption, is discussed, which would allow the turbine to temporarily provide a controllable minimum load to conventional power plants until a sufficient number of consumers has been reconnected. The control system has been implemented and verified using two experimental power converters, each linked to a hardware‐in‐the‐loop (HiL) simulator of a WT and connected to a real medium‐voltage laboratory grid.     

Simulation comparison of wake mitigation control strategies for a two‐turbine case

Wind turbines arranged in a wind plant impact each other through their wakes. Wind plant control is an active research field that attempts to improve wind plant performance by coordinating control of individual turbines to take into account these turbine–wake interactions. In this paper, high‐fidelity simulations of a two‐turbine fully waked scenario are used to investigate several wake mitigation strategies, including modification of yaw and tilt angles of an upstream turbine to induce wake skew, as well as repositioning of the downstream turbine. The simulation results are compared through change relative to a baseline operation in terms of overall power capture and loading on the upstream and downstream turbine. Results demonstrated improved power production for all methods. Analysis of control options, including individual pitch control, shows potential to minimize the increase of, or even reduce, turbine loads.Copyright © 2014 John Wiley & Sons, Ltd.