A review on wind turbine control and its associated methods

Journal of Cleaner Production 174 (2018) 945e953

Contents lists available at ScienceDirect

Journal of Cleaner Production journal homepage: www.elsevier.com/locate/jclepro

Review

A review on wind turbine control and its associated methods
Novaes Menezes, Alex Maurício Araújo*, Eduardo Jose ge Sophie Bouchonneau da Silva Nade UFPE e Federal University of Pernambuco, Department of Mechanical Engineering, Recife, Brazil

a r t i c l e i n f o

a b s t r a c t

Article history: Received 1 August 2017 Received in revised form 1 October 2017 Accepted 27 October 2017 Available online 27 October 2017

Nowadays, the increasing environmental issues especially concerning the global warming have motivated a run for the use of renewable energy sources. Wind energy represents a major player in this context and today it is the most widespread renewable fuel, but still requires many technological improvements. The control of wind turbines (WTs) plays a key role in wind energy applications, ensuring their high efficiency and cost-effectiveness. This has been an intensively researched subject and its developments are crucial to design even better and more efficient wind turbines. However, currently very little papers are addressed to summarize and list wind turbine control concepts. In the present paper, a literature review of wind turbine control is presented dealing with the main wind energy control methods. The main objective of the paper is to form a detailed background to serve as a starting point for new researches on WT control that can be decisive to energetic sustainability. Further, the paper discusses the most recent control developments and their contributions to mitigate environmental issues. © 2017 Elsevier Ltd. All rights reserved.

Keywords: Wind turbine Control MPPT strategies Pitch control Grid integration

Contents 1. 2. 3. 4.

5.

6. 7.

8.

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 945 Methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 946 Control objectives and operational regions of WTs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 946 WT generator torque control and MPPT strategies . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947 4.1. Optimal torque control (OTC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947 4.2. Power signal feedback (PSF) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 947 4.3. Hill-climb search (HCS) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 4.4. Sliding mode control (SMC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 WT pitch control . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 948 5.1. Collective pitch control (CPC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949 5.2. Individual pitch control (IPC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 949 Grid integration control for frequency regulation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950 Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950 7.1. Recent control developments in WT control: LIDAR technology and Model Predictive Control (MPC) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 950 7.2. Recent control developments in WT control: smart rotor applications . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951 7.3. General discussion and the role of wind turbine control in a sustainable energy future . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951 Conclusions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 951 References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . .. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 952

* Corresponding author. UFPE e Federal University of Pernambuco, Department of Mechanical Engineering, Fluid Mechanics Laboratory, Av. da Arquitetura - Cidade
ria, Recife, PE, 50740-550, Brazil. Universita E-mail addresses: [email protected] (E.J. Novaes Menezes), [email protected] (A.M. Araújo), [email protected] (N.S. Bouchonneau da Silva). https://doi.org/10.1016/j.jclepro.2017.10.297 0959-6526/© 2017 Elsevier Ltd. All rights reserved.

1. Introduction Wind is a renewable, clean and endless energy resource, which makes it very suitable to satisfy the increasing energy demand of

946

E.J. Novaes Menezes et al. / Journal of Cleaner Production 174 (2018) 945e953

many countries. Thereby, wind energy has experimented a strong growth worldwide, especially in the last two decades, and the total world installed capacity by the wind industry has changed from 1.29 GW in 1995 to 370 GW by the end of 2015 (GWEC, 2016). Due to environmental concerns and the continuous pursuit for energy security, this scenario should even increase in the next years, with more installed capacity and larger wind turbines (WTs). These machines have been evolving from simple designs towards complex multi-MW generation units, installed together in large arrays named “wind farms”. The complexity of modern WTs forces the control systems to be key components of a wind turbine to ensure safe and efficient operation of these sophisticated wind energy conversion systems (Manwell et al., 2010). This happens because unlike other energy sources, the wind is not controllable. The wind flux is a strongly random process, variable both in time and in space. This variability leads to a difficult conversion of energy, as WTs are subjected to a non-uniform and transient resource, variable mechanical loads and non-linear dynamics. This is the why control systems are indeed very important for WTs. They make possible to cope with the wind variability to produce energy in a reliable and cost-effective manner. The main objectives of the control systems embedded in WTs are maximizing power production, mitigating dynamic and static mechanical loads and guaranteeing a continuous power supply to the grid, according to the utilities requirements. For achieving these goals, the WTs should have operational control systems dedicated to regulate turbine parameters to the desired set-points (Lubosny et al., 2007). The blade pitch angle and the generator torque are the major parameters to be controlled in WTs. Pitch angle control allows to control the wind input torque, in order to enable a smooth power production and to reduce the mechanical loads. On the other hand, the generator torque control allows to vary the WT rotor speed following a Maximum Power Point Tracking (MPPT) strategy, for extracting as much power as possible from the wind flux. Furthermore, the WTs should also have a control of grid integration to control the power delivered to the grid, to provide a wellconditioned electrical power supply. This is necessary because grid integration of WTs becomes a complex task due to the random nature of the wind, which can cause problems to grid frequency stability (Yingcheng and Nengling, 2012). Each one of the above cited control systems (pitch control, generator torque control and grid integration control) has its own technology and methods to be realized, which depend on the WT operational regimes and their respective control objectives. Although there are some specific pitch control and MPPT torque control (Abdullah et al., 2012) and grid integration reviews in the literature (Yingcheng and Nengling, 2011), review papers dealing with the main WT control issues simultaneously are very little reported. Addressing methods for MPPT torque control, WT pitch control and grid integration control in a same paper must be accomplished to construct a general background of invaluable use for future research developments. With the aim of bridging this gap, this review paper presents a literature review of WT control dealing simultaneously with the main WT control systems. The objective consists in making easier the development of future researches on WT control in order to contribute to a more sustainable electrical power generation. Further, once the WT control systems have been analyzed, the paper concludes with a discussion regarding the WT control technology, pointing the future trends that should be applied to improve the wind turbine efficiency, reliability, cost-effectiveness and grid integration. These are mandatory issues to strengthen the wind energy as a renewable and clean energy source.

2. Methods A systematic search of scientific literature was carried out to cover all the aspects and research trends of wind turbine control systems. The main information sources were online scientific databases, even though some classical wind energy books were also consulted to provide a general overview. The online checked scientific databases included the Science Direct, Research Gate and mainly the CAPES Journal Portal, the official scientific search engine of the Brazilian government. The CAPES Journal Portal provides access to more of 37.000 peer-reviewed scientific journals, 66 research bases of thesis and dissertations, 11 research bases of patents and 31 research bases of e-books. The inserted keywords include “wind turbine control”, “control of wind turbines”, “MPPT methods”, “MPPT strategies”, “pitch control”, “wind turbine pitch control” and “grid integration of wind turbines”. In a such vast universe of research, an enormous number of papers were found. Some general criteria have been adopted for each first search with a specific keyword, in order to filter the preliminary results. They are: i) papers must be published only in English, in peer-reviewed journals or in recognized high-quality and traditional conferences about wind energy; ii) papers must be focused specifically on wind turbines and general papers about control must be discarded; iii) papers must deal with one of the following WT control issues: MPPT strategies, pitch control and grid integration. The research has resulted a total of 262 selected papers, including many different WT control issues. Due to space constraints, this number must be reduced to environ 100 papers. To select these, all the papers were analyzed and read carefully to choose the ones most relevant. Afterwards, papers were classified in three categories, according to the main control problem approached: i) WT generator torque control and MPPT strategies; ii) WT pitch control; iii) Grid integration control for frequency regulation. Each category is approached in a specific section of this paper.

3. Control objectives and operational regions of WTs The control objectives of WTs determine the moment of operation of each WT control system and must be clearly established to avoid misunderstanding in analyzing the WT control methods. The definition of control objectives is dependent on the wind turbine operational regions. These are closely related to the wind speed and one can identify three operational regions according to the wind speed (Burton et al., 2011). In the so-called Region 1, below the cutin wind speed, there is no production of electrical power, as the wind speed is too low and the produced power would not compensate the losses in the turbine operation. In this operational region, the turbine should be stopped or in idle mode. Region 2, between the cut-in and the rated wind speed, has an increasing power production as the wind increases progressively. In Region 2 the WT is in the partial-load regime. In Region 3, wind reaches rated speed and WT enters in full-load regime. Power production must be limited to the WT rated power, for ensuring operation within the safety limits of generator speed and WT mechanical loads. Some authors also identify a Region 2.5, where the WT rotor has achieved rated speed but the torque is still below its rated value (Aho et al., 2013). One can also consider the additional Region 4, after the cut-out wind speed, where the WT must be switched off due to the very high wind regimes. The power extracted from the wind can be expressed according to Eq. (1) (Rohatgi and Vaughn, 1994):

E.J. Novaes Menezes et al. / Journal of Cleaner Production 174 (2018) 945e953

Pw ¼

1 rACp v3 2

(1)

where Pw is the power extracted from the wind, r is the air density, A is the rotor swept area, v is the wind speed and Cp is the so-called power coefficient, which depends on the pitch angle b and on the tip speed ratio l. Tip speed ratio is defined by the relation between linear velocity on the tip blade and wind speed, according to Eq. (2):

l¼

UR v

(2)

where U is the rotor speed and R is the rotor radius. Field tests and simulations have shown that the power coefficient Cp ðl; bÞ is maximum for a predetermined optimal pitch angle (bopt ) and optimal tip speed ratio (lopt ), which are specific constants for each particular WT. When operating in Region 2 the WT rotor speed should be varied to maintain this optimal tip speed ratio (TSR) as the wind changes its speed, for ensuring the maximum power production. The way the rotor speed is varied is the Maximum Power Point Tracking (MPPT) strategy. In Region 3, the control objective changes from maximizing to limiting power, as well as limiting rotor speed and torque. The generator torque is maintained constant at its rated value and the pitch angle should be controlled to reduce the power coefficient and aerodynamic efficiency. Thereby, constant rated power is extracted from the wind. One can summarize control objectives for the operational regions of WTs as follows:  Regions 1 and 4: The WT should be out of operation, commanded by the supervisory control.  Region 2: Maximize power production via MPPT strategies and generator torque control, which is the control system typically active in this region; the generator torque control should be a trade-off between generator torque actuation and optimal power production.  Region 2.5: The rated speed should be maintained constant and the torque should be slightly increased until its rated value, ensuring a smooth transition between Regions 2 and 3.  Region 3: Power production should be limited to the rated power via pitch control, which is the control system typically active in this region. Additionally, as the wind speed is above rated, the control objective of mechanical load reduction becomes important due to the high wind speeds that can damage the WT structure.

4. WT generator torque control and MPPT strategies The generator torque control sets the WT rotor speed in order to obtain maximum power production, via variable speed operation. The controller must adjust the generator torque for accelerating or decelerating the turbine (Manwell et al., 2010). The WT can be modeled as a system subjected to both wind torque and generator torque and with inertia, damping and stiffness. As the wind speed changes, wind torque increases or decreases and the generator torque must be the actuator for taking the dynamic system to the optimal operating point. The early WTs had not torque control. WTs used Squirrel Cage Induction Generator (SCIG) connected directly to the grid, thus the generator torque was not controllable and the generator speed was tied to the grid frequency. The possibility of variable speed operation emerged due to the advances in solidstate devices. Electronic power converters (PCs) became suitable to be used in WTs. The PCs are devices that interface the generator

947

with the grid, by decoupling the generator speed from the grid frequency. PCs execute the electronic switching of stator and rotor voltages to produce the desired voltage frequencies, making possible to control generator torque and the active/reactive power produced in the generator. Nowadays, most of the generators used in variable-speed operation include the Doubly-Fed Induction Generator (DFIG) and the Permanent Magnet Synchronous Generator (PMSG). The WT torque control is realized by a two-layer controller in a cascaded structure (Rajendran and Jena, 2014). The external control loop determines the torque set-point via MPPT strategy. The inner control loop, also called the electrical control loop, is implemented via PCs and receives the torque set-point from the external control loop. Since the generator electrical dynamics is much faster than the WT mechanical dynamics, the electrical control loop is considered to be an actuator and to have no significative control concerns. Control must be focused on the implementation of MPPT strategy. The torque set-point should take the WT to the optimal operation point while maintaining reduced loads. To execute the task, there are many MPPT strategies proposed in the literature. In the following, a review of the researches on the existing MPPT methods and modern control techniques used in MPPT will be presented. 4.1. Optimal torque control (OTC) The optimal torque control is the standard MPPT algorithm that is largely implemented in commercial wind turbines. It relies on imposing the torque according to a quadratic law based on the rotor speed, Eq. (4):

Tg ¼ kU2 k¼

C 1 rpR5 p;opt 2 l3

(4)

opt

where k is a constant derived from the aerodynamic characteristics for each particular WT, calculated according to the optimal values to the power coefficient Cp;opt and the tip speed ratio lopt . OTC is a very simple method and ensures operation around the optimal power coefficient (Manonmani and Kausalyadevi, 2014). As mitigating mechanical loads is expected, it may include an additional torque control loop to damp drive-train torsion mode and resonant loads. OTC requires only a speed shaft sensor, which is normally available in commercial WTs. It is a classical method and it is frequently used as baseline for comparison with advanced controllers in research papers. Despite its advantageous simplicity, OTC has some important drawbacks. The precise value of k is not easy to obtain and needs to be calculated for each WT. Moreover, the blades’ aerodynamic conditions change over time due to wear or dirt, and hence the value of k can change during turbine operation. High wind turbulence can also lead to a poor performance, as the WT will not operate in steady conditions during much time and OTC is essentially a steady-state method (Yu Zou et al., 2013). Some research suggest adaptive control approaches to face parameters uncertainty, using an adaptive value for k to obtain better performance (Johnson, 2008). 4.2. Power signal feedback (PSF) This MPPT strategy supposes a power closed-loop control. The control law follows essentially the same operational trajectory as the optimal torque control, but uses a power control loop rather than directly imposing the torque (Cheng and Zhu, 2014). The technique has the same disadvantages of OTC, namely the need for a precise value of the optimal power coefficient and optimal tip

948

E.J. Novaes Menezes et al. / Journal of Cleaner Production 174 (2018) 945e953

speed ratio as well as the poor performance at turbulent winds. In early studies as (Muljadi, 2001), it is established that the smoothness in rotor speed variations depends on the difference between the electrical power reference and the mechanical wind power. Furthermore, using PSF makes possible adjusting active and reactive power set-points, and thus a better grid integration (Hansen et al., 2005). The power control loop can be established in different ways. Hilloowala and Sharaf (1994) uses a PI-controller (proportional-integral) to regulate the output power. The power reference is taken from wind speed and not from shaft speed as in the standard PSF. In Kim et al. (2015), the gains for PI-controller are optimized using Particle Swarm Optimization (PSO). Gainscheduled control is considered in Bianchi et al. (2012), where a multivariable problem is solved using H∞ techniques. A general procedure for gain-scheduling is developed and authors state that any linear control method can be used according to the presented scheduling procedure. Despite the diversity of implementations in PSF, all of them focus on setting the active power reference according to the optimal wind power available. 4.3. Hill-climb search (HCS) This method is quite robust and does not depend on a priori knowledge of WT’s characteristics. Neither optimal power coefficient nor optimal TSR needs to be known, as they do in PSF and OTC control. HCS is an iterative procedure where the controller sends a signal of speed variation (DU) to the WT and observes the resulting power variation (DP). Thereafter, the sign of the ratio DP=DU determines if the speed should be increased or reduced (Raza Kazmi et al., 2011). This is possible because the relation between tracked power and rotor speed has a parabolic allure, characterized by a single point of maximum for each wind speed. This method is mainly used for low and medium scale WT. Large scale WT has a large rotor inertia and changes in speed or power are more difficult to arise and be observed. Due to its characteristics, this method is
lez et al., 2010). also known as Perturb and Observe (P&O) (Gonza There are many possibilities for implementing HCS. The main concern is how the perturbation should be applied. A weak perturbation can lead to a slow response, while a strong one can stress excessively the WT or even make the system unstable (Abdullah et al., 2012). Various papers use different techniques to apply the perturbation and take decisions based on the observed variable. In an early paper Yaoqin et al. (2002) uses the HCS method with a variable step perturbation, proportional to the derivative dP=dU of the power curve. This is supposed to increase the convergence of HCS and to lead to a strong improvement in energy capture. A fixed step perturbation was used by Datta and Ranganathan (2003), where the speed reference is set proportionally to the speed perturbation. An experimental test shows that the employed method performs as well as the traditional OTC. Fuzzy-based and Neural Networks techniques can also be employed for obtaining speed references in HCS method, e.g. in Li et al. (2006). More recent work considers a synergy between HCS and PSF techniques. In Zhu et al. (2012) the HCS method is utilized for determining a power reference, which will be the input for a power control loop. A comparison is made with TSR control and authors establish this hybrid approach can obtain a good power tracking. 4.4. Sliding mode control (SMC) Sliding mode control is a modern and non-linear control method recently researched for WT applications (Oudah et al., 2014). SMC is especially suited for WT due to its recognized

robustness to parameters uncertainties and non-linear structure. The method relies on desired control actions that are switched suitably, leading to the term “sliding”. As a general closed-loop technique, it can be applied for TSR control or PSF control. One of the earlier papers proposing SMC for WT was Battista et al. (2000), where the SMC is designed for power maximization and the robustness of the method is highlighted. Since them, much research effort has been done and the most recent papers deal with sliding mode control with the aim of maximizing energy capture whilst reducing drive train loads. In this context, research in Munteanu et al. (2008) considers the sliding surface as dependent on a variable representing the desired trade-off between torque effort and energy capture. Two different strategies for SMC were proposed by Aguilar and Jorge (2014). Authors design first-order and second-order sliding mode controllers for tracking the optimal tip speed ratio. Next, a high-order quasi-continuous sliding mode controller was established based on power signal feedback. Both strategies are compared with standard OTC control and simulations show better results in terms of power capture and dynamic characteristics. Higher order SMCs are frequently pointed as capable of reducing chattering phenomenon, which represents less drive-train vibrations (Evangelista et al., 2010). Applications of sliding control using PSF can also be considered, where a SMC controller obtains reduction in drive train mechanical stresses (Beltran et al., 2008). Some research proposes SMC with an integral control action imposed to the sliding surface. Saravanakumar and Jena (2015) propose a SMC and an integral sliding mode controller (ISMC). An important advantage of ISMC is the continuity of the control action between Regions 2 and 2.5 that can afford better control transition.

5. WT pitch control Pitch control allows to change the pitch angle of WT’s blades in order to control its aerodynamic efficiency. The pitch angle is a major WT parameter as it determines the wind’s angle of attack (Hau and von Renouard, 2005). Thus, turning the blades around their own axes changes the relative wind flow and consequently the aerodynamic loads exerted on the rotor. Moreover, the power coefficient Cp ðl; bÞ varies according to the pitch angle and consequently the power capture varies as well. Therefore, pitch angle control has a twofold role: power regulation and load reduction. These are prominent in operation on Region 3, where the power production must be limited to the rated one and the high wind speeds imposes severe loads to WT structure and rotor (Johnson et al., 2011). Different concepts have been proposed along the years for power regulation and load reduction of WTs. The early WTs were controlled by passive means using the blades’ aerodynamic characteristics. The airfoils were designed to stall when subjected to high wind regimes, which is called passive stall control. None additional actuator was necessary, thus a simple and low-cost power control was realized. However, the controllability was very limited as it was based on a natural stall phenomenon without any active control. In passive stall control, the WT is subjected to more power fluctuations, torque spikes and varying load effort. For overcoming these drawbacks, modern WTs use active pitch control, with electrical or hydraulic pitch actuators (Chiang, 2011). In Region 3, the generator torque is usually imposed to be constant while the pitch control should be working to maintain constant rotor speed. While changing the pitch angle the control system is also changing the wind torque and accelerating or decelerating the turbine. As torque and speed are set to be constant, the power production is limited and mechanical loads are reduced, reaching

E.J. Novaes Menezes et al. / Journal of Cleaner Production 174 (2018) 945e953

the control objectives of Region 3. Typically, the pitch control loop uses only the rotor speed as feedback signal and the pitch commanded value is the same for the three blades. This is the so-called collective pitch control (CPC). A modern researched pitch control with the specific aim of load reduction is the method of individual pitch control (IPC). Both CPC and IPC techniques are reviewed in the following.

949

baseline CPC feedback control. A more recent research conducted by Koerber and King (2013), suggests a Model Predictive Control (MPC) designed using LIDAR measurements as preview information. MPC is by definition a multivariable and constraint handling method. Authors use torque and pitch as control signals and consider constraints on pitch rate. MPC allows an optimization procedure and the simulations show better performance than the traditional PID controller. Finally, field tests proving effectiveness of feedforward CPC can be found in Schlipf et al. (2014).

5.1. Collective pitch control (CPC) This is the traditional method of pitch control, which is largely € ffker, 2016). The implemented in commercial turbines (Njiri and So commanded pitch is sent collectively for the blades, meaning that the same control action is taken in each one of them. Usually, CPC implementation relies on a simple PID (proportional-integral-derivative) control law, with rotor speed variation being the error signal for the closed-loop control. A systematic simulation-based procedure for selecting the PID control gains can be found in Hand and Balas (2002). As the WT is a nonlinear system, applying a linear control law such as this simple PID requires linearization around an operating point. Different operational conditions degrade controller’s performance and hence gain scheduling techniques are necessary for improving efficiency (Jonkman et al., 2009). Such techniques are widely present in the literature (Bianchi et al., 2006; Li et al., 2015). Although PID control with gain scheduling is the classical CPC implementation, the constant pursuit for load reduction has motivated the rise of modern CPC approaches in many research works. Robust and adaptive methods are used to overcome modelling uncertainties. Frost et al. (2009) develop adaptive pitch control with concerns to disturbance rejection. This work is followed by Frost et al. (2011), where this adaptive controller is extended with a Residual Mode Filter (RMF) to avoid modal WT subsystems that can be excited during operation especially in turbulent conditions. On the other hand, nonlinear modelling is used in Bououden et al. (2012), where fuzzy methods develop a T-S (Takagi-Sugeno) model for pitch control purposes. A hybrid Fuzzy-PI control is considered in Duong et al. (2014), where the focus is CPC to smooth power fluctuations. Controllers based on neural networks, although less present in the literature, can also be envisaged (Poultangari et al., 2012; Yao et al., 2010). CPC with fault tolerant control (FTC) has been researched for installations of hard maintenance, as offshore WT. Fault tolerant capability is essential to reduce WT downtimes and fault detection and FTC are pointed as critical research fields for the wind industry (Laks et al., 2009). In Sloth et al. (2011), authors propose both active and passive fault tolerant pitch controllers, which can perform well even in the presence of leakage in the pitch hydraulic system. The active FTC (AFTC) relies on information from a fault diagnosis system while passive FTC (PFTC) does not depend on such a system, but is less performant. A more sophisticated approach is used by Luzar and Witczak (2014). System identification is proceeded by neural network for assembling a Linear Parameter Varying (LPV) system, which allows handling system nonlinearities and develop an AFTC. Another promising field of research is predictive and feedforward CPC, due to the advances in LIDAR (Light Detection and Ranging) applications. LIDAR is a sensing technology used to remotely measure the wind speed based on laser diffraction. An early work considering feedforward-CPC has simulation proven significant load reductions, in order of 10% (Harris et al., 2006). Since them, many papers have demonstrated such benefits. Dunne et al. (2011) considers LIDAR measurements to design a feedforward controller using a non-causal series expansion for inverse-model control. Inverse-models aim on cancelling disturbances affecting controlled outputs. The proposed controller has performed better than

5.2. Individual pitch control (IPC) IPC is the most recent development in pitch control, which has been intensively researched in the last years, but still not completely implemented in commercial WTs. It is expected to be largely applied in the next generation of turbines to shift WT’s design for more and more larger and flexible blades (Shan et al., 2013). IPC is a technique frequently pointed in the literature as capable of reducing loads and fatigue damage (Bossanyi, 2003; Geyler and Caselitz, 2007; Petrovi
c et al., 2015). It makes the WT control system an inherently Multiple-Input-Multiple-Output (MIMO) system, since it requires individual pitch commands for each blade and the presence of additional sensors. These could be strain gauges or accelerometers enabling measuring of variables such as blade root moment or tower displacement. Based on the additional measures, the controller performs an additional control action that is individual to each blade. The aim is to adjust the pitch angle to reduce the blade root moment or damping structural modes. Individual pitch command works in a different frequency range from the collective pitch one. The pure CPC has the objective of regulating rotor speed while IPC envisages little adjustments in pitch angle for reducing stresses. Although simulations prove the benefits of using IPC, the field tests and practical implementation remain in course (van Solingen et al., 2015). In the point of view of hardware, the challenge remains about sensors reliability, as the modern WTs are already equipped with individual pitch actuators for each blade. Sensors installed at the blades would operate under several conditions and difficult maintenance (Ehlers et al., 2007). There is also a concern about the wear in individual pitch mechanisms due to the increased actuation cycle (Jelavi
c et al., 2010). Despite these practical questions, IPC papers generally focus on control methods and this is a fertile research field. Early proposals of IPC relied on feedback control based on structural sensors (Bossanyi, 2005). The most recent development consists in using LIDAR technology for feedforward control. The preview information about wind inflow provided by LIDAR allows the control system to act in advance to incoming wind events, as gusts or disturbances. Laks et al. (2011) compare feedforward IPC to the IPC feedback only, based on H∞ techniques. Simulations results show reductions in damage equivalent loads (DEL), a standard measure of fatigue damage. A technical report from the U.S. National Renewable Energy Laboratory (NREL) deals with different LIDAR-based strategies for IPC, comparing preview time and LIDAR implementations (Dunne et al., 2012). Feedforward strategies using MPC, which is a MIMO method in nature, are also very suitable for IPC. In an early study conducted by Henriksen (2007) a linear MPC controller is proposed, while Kumar and Stol (2009) suggest a scheduling procedure, which allows improved performance due to the most refined tuning. LIDAR preview along with local inflow measurements are used to develop a more precise MPC controller in the work of Kragh and Hansen (2010). Short-term wind field predictions can also be used for MPC purposes and are considered in a recent research (Spencer et al., 2013). At last, one can cite hybrid methods as the most recent advance in MPC pitch control,

950

E.J. Novaes Menezes et al. / Journal of Cleaner Production 174 (2018) 945e953

presented in Navalkar et al. (2014). In this research, authors consider a repetitive control (RC), which is a modern ‘learning’ control technique, helped by MPC for constraint handling goals. Another focus of research is related to IPC applications in offshore WTs. Researchers have considered the possibility of using individual pitch to help in stabilization of proposed floating WT platforms. IPC based on disturbance accommodating control can be envisaged for barge and tension leg platforms (Namik and Stol, 2011). MPC is also a feasible alternative in offshore WTs, developed for reducing blade loads and yaw rolling of a barge platform (Chaaban and Fritzen, 2014). Besides offshore applications, studies of IPC effects on ultimate loads affecting the WT are present. In fact, the traditional IPC aims to mitigate fatigue loads, but it can also be profitable for reducing ultimate loads (Bottasso et al., 2014a; Corcuera et al., 2014). Finally, confirming IPC as a future trend for larger WTs, the work of Chen and Stol (2014) and references therein, considers IPC performance in very large WTs with rating up to 15 MW, through an upscaled model, obtaining good results. On the other hand, the most recent field tests in normal scale were performed by Bossanyi et al. (2013) on NREL’s experimental WT CART3 (3-bladed Control Advanced Research Turbine). The tests show once again an effective load reduction, as it was expected. 6. Grid integration control for frequency regulation The integration of WT into grid power systems presents important challenges due to the characteristics of wind generation. The randomness and fluctuating nature of the wind can cause problems for both power system stability and power quality (Saqib and Saleem, 2015). Power quality refers to concerns such as flicker and harmonics, which implies controlling the voltage within a strict range of values. On the other hand, power system stability encompasses frequency regulation, which requires maintaining the power system’s nominal frequency regardless the variations of electrical load demand. Thereby, grid integration control of WT is a twofold problem. Regarding to power quality, much research has been done and advances in power electronics are improving WT power quality using reactive power compensation and voltage control (Girbau-Llistuella et al., 2014; Huang et al., 2015). Frequency regulation requirements emerge as the wind power becomes increasingly widespread. It is easy to prove that problems related to frequency stability may arise due to increasing participation of WT in power systems (B. Jain et al., 2015b). In the last years, with the strong growth of wind power, frequency regulation has become a main concern in WT grid integration. It is also the most recent development in WT control. Many WT manufacturers have recently reached the minimal requirements of frequency regulation capabilities, but research is still ongoing to understand the limits and possibilities of this technology (Marschner et al., 2014). WTs are different from conventional power plants, where the fuel is controllable and can be adjusted according to the frequency needs. The main objective of a power system is to provide electrical power guaranteeing the balance between power demand and power supply. If there is an imbalance, the frequency will suffer a drop, when the demand is larger than the supply, or an increase, when the supply is larger than the demand. Alterations on frequency can damage devices connected to the grid and can even lead to a grid fault with system blackout. Therefore, an increased level of wind energy penetration must be accompanied by frequency regulation capability of WT and some utilities grid codes already impose it (Vogler-Finck and Früh, 2015). One can classify the frequency adjustments provided by WT according to different time scales. Inertia emulation is the shorter one and is used to slow down the initial rate of change in frequency. Primary frequency control (PFC) is the next and is used to bring frequency to a steady-

level. The last one corresponds to automatic generation control (AGC), used to bring frequency back to its nominal level (Ela et al., 2014). As frequency depends on balancing load and supply, the electrical power generated by WT should be controlled to certain values. This is the so-called active power control (Wang and Seiler, 2014). It could be implemented using both torque or pitch controllers on different WT operational regions. According to these, there are many research papers developing WT frequency regulation capabilities. The early proposals for frequency regulation were made by Rodriguez-Amenedo et al. (2002), where pitch control was used to follow a variable power reference and thus implement AGC. Torque control for the same purposes was considered in Pudjianto et al. (2007). More recently, the torque methods have been considered for de-rated operation, which consists in maintaining a power reserve that can be used for PFC purposes. However, modern multivariable control strategies are the newest approaches for PFC. In Camblong et al. (2014), multivariable control is developed for frequency regulation via a MIMO LQG controller, which considers also individual pitch and torque controllers acting simultaneously in full-load region. MPC is another multivariable control used for frequency regulation, in different contexts. In Kassem (2012), pitch control based on MPC is considered for load-frequency adjustments. A frequency control for a multi-area power system using MPC is developed in (Mohamed et al., 2012). Hybrid power systems e.g. wind-diesel can also be frequency-controlled by MPC (Kassem and Yousef, 2013). Despite the above modern methods for PFC and AGC, inertia emulation has been traditionally studied according to simple proportional-derivative control (Attya and Hartkopf, 2013; Miller et al., 2012). However, a recent research presents an optimization procedure based on particle swarm optimization (PSO) (Hafiz and Abdennour, 2015). To sum up the above papers, one can state that, although the multiple focus of research, the main strategy is to develop frequency control by operating WT in a speed higher than the optimal, as shown in the complete report by Ela et al. (2014). Due to this non-optimal operation, there is a power reserve that can be enabled changing the operational point to the optimal one. The power reserve is used when necessary to balance frequency drops. At last, beyond the research works, it is remarkable the effort of wind industry R&D for achieving high levels of WT’s frequency reliability. The technology is proprietary but it is worthy to note that patents are already being registered (Clark et al., 2010). 7. Discussions 7.1. Recent control developments in WT control: LIDAR technology and Model Predictive Control (MPC) LIDAR (Light Detection and Ranging) is a sensing technology recently researched for WT control applications. It consists in a powerful tool for determining the wind field some meters ahead the WT rotor. LIDAR emits a coherent light beam that is backscattered by natural aerosols such as particles of dust and water droplets. The reflected light is detected and processed according to Doppler effect shifts in order to determine the wind speed. This technology paves the way for new control concepts in wind energy such as feedforward control and Model Predictive Control (MPC). The main LIDAR-based control method is the MPC. This is a modern control strategy, which considers future states as inputs for optimizing the control law (Wang, 2009). LIDAR can provide these future states by measuring the wind speed ahead of the rotor. MPC is a multivariable control and research papers consider both torque and pitch as control inputs, to achieve a single control framework for Regions 2 and 3. Bottasso et al. (2014b) envisages two MPC

E.J. Novaes Menezes et al. / Journal of Cleaner Production 174 (2018) 945e953

controllers, Receding Horizon Control (RHC) and a predictive LQR. LIDAR is used to predict the wind speeds. Trial and error procedure determines the optimal horizon of prediction. Simulations results show that predictive LQR outperforms the traditional LQR control, mainly in turbulent wind regimes. Jain et al. (2015a) uses a systematic procedure for tuning a MPC controller according to optimization criteria such as average power output and drive train twist rate. Further, there are MPC papers in which wind preview is obtained by estimation techniques instead of LIDAR measures. For example, in Kusiak et al. (2010) a time series is used to estimate the wind speed and implement a MPC controller. Authors also use a data mining procedure for WT modelling, extracting data from the SCADA system. An important characteristic of MPC is its ability to handle constraints. Torque actuation rate or rotor speed limits are constraints that can be explicitly taken into account in MPC formulation. A constraint-handling MPC is studied in Henriksen et al. (2012), where a direct comparison is made with PI controllers. Authors demonstrate the superiority of the proposed approach in attending constraints of maximum generator power. At last, the most recent trend in MPC is the robust MPC (RMPC), which considers uncertainties of the variables on the prediction horizon (Mazenc et al., 2015). Beyond the modern MPC methods, applications of LIDAR to traditional control shows up in (Wang et al., 2013), where Optimal Torque Control (see Sec. 4.1) using LIDAR measures is compared to a feedforward torque controller. It is demonstrated by simulations that the latter has the better performance considering load reduction and power capture. At last, it is also possible to include LIDAR-based control for enhancing energy capture in Region 2.5 by developing feedforward torque controllers (Wang et al., 2014). 7.2. Recent control developments in WT control: smart rotor applications The smart rotor is an intensive field of research dealing with load reduction for the future WTs. The smart concept includes the presence of local sensors and actuators distributed along the blades with embedded intelligence. It is a new approach of active load control of WTs, strongly researched in the last decade. Rather than controlling the pitch angle by turning the entire blade commanded by a pitch motor, the smart rotor controls the relative wind flow or loads by using specific actuators located spanwise along the blade. This makes the WT rotor a smart structure, capable of reacting faster and precisely to load events, such as wind gusts. The early investigations on smart blades were developed by NREL in the middle 1990s. At that time, integrated blade flaps for aerodynamic braking have been considered (Miller, 1995; Quandt and Migliore, 1996). Thereafter many proposals for installing different active load control devices in WT blades were made. Possible technologies are trailing-edge flaps (Bergami and Gaunaa, 2010; Smit et al., 2015), microtabs (Dam et al., 2007; Nakafuji et al., 2001), active blade twist (Kota, 2008), and devices for boundary layer control, e.g. plasma actuators or synthetic jets (Jukes, 2015). Simulations have been showing promising results for smart rotor applications (Vries et al., 2014). Modern control techniques are employed, such as MPC controllers, H∞ control and iterative learning control. Wind tunnels experiments in reduced-scale models have also been realized. However, field tests are still at the beginning. Currently, there are just two field tests presented in the literature, both using flaps as active devices in Castaignet et al. (2014) and (Berg et al., 2014). Implementation of smart rotors on commercial WTs presents some difficulties. First, the actuators and sensors to be installed on the blades should reach a high level of reliability, due to the hard maintenance and to avoid increasing WT operational costs. Moreover, implementing these devices would require an

951

expensive redesign of the blades. For these reasons, some authors consider that smart concepts would work as auxiliary rather than the only load control systems in the future WTs. In this context, smart rotor could act to complement pitch control systems (Plumley et al., 2014). 7.3. General discussion and the role of wind turbine control in a sustainable energy future The main WT control concerns, namely generator torque control, pitch control and grid integration control, are interconnected and should contribute together for an optimal operation of WT. Many countries have established ambitious goals for wind power implementation in the next future. This should require a WT shift to a higher performance level and increased power ratings. As the complexity and size of these machines increase, the control is expected to assume an increasing role in the future WTs. Control developments are fairly necessary and so the control research must be strongly ongoing. The control can contribute for a sustainable energy future in three main branches: improving WT efficiency, reducing WT mechanical loads to allow more flexible and cheap components and reducing WT mechanical loads to allow increased WT lifetimes. These three points acting together can reduce the cost of energy (COE) of wind energy, making it more competitive regarding polluting energy sources, such as fossil fuels. In fact, nowadays the main obstacle to the wind energy development is its relatively high COE compared to fossil fuels, e.g. coal and natural gas. Thus, it is essential to continue control research and advance in control methods and technologies. The new sensing technologies as LIDAR are not fully explored yet. LIDAR enables the effective deployment of feedforward control, which is not currently implemented in commercial WTs. More research about feedforward methods needs to be made, as in Model Predictive Control and combined feedback-feedforward control. Furthermore, the uncertainties of LIDAR measurements need to be better understood and taken into account. Individual Pitch Control is another technology still not totally implemented and needing further research. Future developments should include more field tests, concerns about reliability of the required new sensors and even more advanced control algorithms. It is also required investigation of the economic benefits carried by IPC, e.g. to quantify increased WT lifetime or reduced cost of the blades, due to the reduced mechanical loads. On the other hand, the smart rotor, which is a more recent approach for load reduction, lies in an earlier development stage and is a completely ongoing field of research. Last but not least, there are important research opportunities for grid frequency regulation. Providing active power control (APC), which enables frequency regulation capabilities, has an impact on WT mechanical loads that rests to be determined. Additionally, many control methods have not been tested yet, as APC has been recently implemented on commercial WTs. New forms of APC, using storage systems, should also be theme of future research. 8. Conclusions This paper has reviewed the control of WT providing a general background and a literature survey about the subject. It has been a comprehensive review since it has considered the three main WT control concerns, namely MPPT methods, pitch control and grid integration control. As shown in the present paper, control of WT is a complex and interdisciplinary subject. Nevertheless, the paper has attempted to give a thorough discussion over the main control systems in WTs. It is expected that this paper may serve as a general reference and a starting point for future researches.

952

E.J. Novaes Menezes et al. / Journal of Cleaner Production 174 (2018) 945e953

References Abdullah, M. a, Yatim, a. H.M., Tan, C.W., Saidur, R., 2012. A review of maximum power point tracking algorithms for wind energy systems. Renew. Sustain. Energy Rev. 16, 3220e3227. https://doi.org/10.1016/j.rser.2012.02.016. Aguilar, L.T., Jorge, D., 2014. Analysis and synthesis of sliding mode control for large scale variable speed wind turbine for power optimization. Renew. Energy 71, 715e728. https://doi.org/10.1016/j.renene.2014.06.030. Aho, J.P., Buckspan, A.D., Dunne, F.M., Pao, L.Y., 2013. Controlling wind energy for utility grid reliability. ASME Mech. Eng. Mag. 145. Attya, A.B., Hartkopf, T., 2013. Wind farms dispatching to manage the activation of frequency support algorithms embedded in connected wind turbines. Int. J. Electr. Power Energy Syst. 53, 923e936. https://doi.org/10.1016/ j.ijepes.2013.06.011. Battista, H. De, Puleston, P.F., Mantz, R.J., Christiansen, C.F., 2000. Sliding mode control of wind energy systems with DOIG d power efficiency and torsional dynamics optimization. IEEE Trans. Power Syst. 15, 728e734. Beltran, B., Ahmed-ali, T., El, M., Benbouzid, H., 2008. Sliding mode power control of variable-speed wind energy conversion systems. IEEE Trans. Energy Convers. 23, 551e558. Berg, J.C., Resor, B., Paquette, J., White, J., 2014. SMART Wind Turbine Rotor: Design and Field Test. SAND2014e0681. Sandia Natl. Lab. Bergami, L., Gaunaa, M., 2010. Stability investigation of an airfoil section with active flap control. Wind Energy 13, 151e166. https://doi.org/10.1002/we.354. Bianchi, F.D., De Battista, H., Mantz, R.J., 2006. Wind Turbine Control Systems: Principles, Modelling and Gain Scheduling Design. Springer-Verlag, New York.
nchez-Pen ~ a, R.S., Guadayol, M., 2012. Gain scheduled control based Bianchi, F.D., Sa on high fidelity local wind turbine models. Renew. Energy 37, 233e240. https:// doi.org/10.1016/j.renene.2011.06.024. Bossanyi, E.A., 2005. Further load reductions with individual pitch control. Wind Energy 8, 481e485. https://doi.org/10.1002/we.166. Bossanyi, E.A., 2003. Wind turbine control for load reduction. Wind Energy 6, 229e244. https://doi.org/10.1002/we.95. Bossanyi, E. a, Fleming, P. a, Wright, A.D., 2013. Validation of individual pitch control by field tests on two-and three-bladed wind turbines. IEEE Trans. Control Syst. Technol. 21, 1067e1078. Bottasso, C.L., Croce, A., Riboldi, C.E.D., Salvetti, M., 2014a. Cyclic pitch control for the reduction of ultimate loads on wind turbines. J. Phys. Conf. Ser. 524, 12063. https://doi.org/10.1016/j.renene.2012.08.079. Bottasso, C.L., Pizzinelli, P., Riboldi, C.E.D., Tasca, L., 2014b. LiDAR-enabled model predictive control of wind turbines with real-time capabilities. Renew. Energy 71, 442e452. https://doi.org/10.1016/j.renene.2014.05.041. Bououden, S., Chadli, M., Filali, S., El Hajjaji, A., 2012. Fuzzy model based multivariable predictive control of a variable speed wind turbine: LMI approach. Renew. Energy 37, 434e439. https://doi.org/10.1016/j.renene.2011.06.025. Burton, T., Jenkins, N., Sharpe, D., Bossanyi, E., 2011. Wind Energy Handbook. Wiley. Camblong, H., Vechiu, I., Etxeberria, A., Martínez V, 2014. Wind turbine mechanical stresses reduction and contribution to frequency regulation. Control Eng. Pract. 30, 140e149. https://doi.org/10.1016/j.conengprac.2014.03.007. Castaignet, D., Barlas, T., Buhl, T., Poulsen, N.K., Wedel-Heinen, J.J., Olesen, N.A., Bak, C., Kim, T., 2014. Full-scale test of trailing edge flaps on a Vestas V27 wind turbine: active load reduction and system identification. Wind Energy 17, 549e564. https://doi.org/10.1002/we.1589. Chaaban, R., Fritzen, C., 2014. Reducing blade fatigue and damping platform motions of floating wind turbines using model predictive control. In: Proceedings of the 9th International Conference on Structural Dynamics EURODYN 2014, pp. 3581e3588. Chen, Z.J., Stol, K. a, 2014. An assessment of the effectiveness of individual pitch control on upscaled wind turbines. J. Phys. Conf. Ser. 524, 12045. https://doi.org/ 10.1088/1742-6596/524/1/012045. Cheng, M., Zhu, Y., 2014. The state of the art of wind energy conversion systems and technologies: a review. Energy Convers. Manag. 88, 332e347. https://doi.org/ 10.1016/j.enconman.2014.08.037. Chiang, M.H., 2011. A novel pitch control system for a wind turbine driven by a variable-speed pump-controlled hydraulic servo system. Mechatronics 21, 753e761. https://doi.org/10.1016/j.mechatronics.2011.01.003. Clark, K., Miller, N., Sanches-Gasca, J., 2010. Modeling of GE Wind Turbinegenerators for Grid Studies. Corcuera, a D. De, Pujana-Arrese, a, Ezquerra, J.M., Segurola, E., Landaluze, J., 2014. Wind turbine load mitigation based on multivariable robust control and blade root sensors. J. Phys. Conf. Ser. 555, 12024. https://doi.org/10.1088/1742-6596/ 555/1/012024. Dam, C.P. Van, Chow, R., Zayas, J.R., Berg, D.E., 2007. Computational investigations of small deploying tabs and flaps for aerodynamic load control. J. Phys. Conf. Ser. 75, 12027. https://doi.org/10.1088/1742-6596/75/1/012027. Datta, R., Ranganathan, V.T., 2003. A method of tracking the peak power points for a variable speed wind energy conversion system. IEEE Trans. Energy Convers. 18, 163e168. https://doi.org/10.1109/TEC.2002.808346. Dunne, F., Pao, L.Y., Wright, A.D., Jonkman, B., Kelley, N., 2011. Adding feedforward blade pitch control to standard feedback controllers for load mitigation in wind turbines. Mechatronics 21, 682e690. https://doi.org/10.1016/ j.mechatronics.2011.02.011. Dunne, F., Schlipf, D., Pao, L.Y., Kelley, N., Wright, A.D., Simley, E., Jonkman, B., 2012. Comparison of two independent lidar-based pitch control designs. In: 50th

AIAA Aerosp. Sci. Meet. Incl. New Horzons Forum Aerosp. Expo., pp. 1e19. Duong, M.Q., Grimaccia, F., Leva, S., Mussetta, M., Ogliari, E., 2014. Pitch angle control using hybrid controller for all operating regions of SCIG wind turbine system. Renew. Energy 70, 197e203. https://doi.org/10.1016/ j.renene.2014.03.072. Ehlers, J., Diop, A., Bindner, H., 2007. Sensor selection and state estimation for wind turbine controls. In: 45th AIAA Aerospace Sciences Meeting and Exhibit, Aerospace Sciences Meetings. American Institute of Aeronautics and Astronautics. https://doi.org/10.2514/6.2007-1019. Ela, E., Gevorgian, V., Fleming, P., Zhang, Y.C., Singh, M., Muljadi, E., Scholbrook, a, Aho, J., Buckspan, a, Pao, L., Singhvi, V., Tuohy, a, Pourbeik, P., Brooks, D., Bhatt, N., 2014. Active Power Controls from Wind Power: Bridging the Gaps. Evangelista, C., Puleston, P., Valenciaga, F., 2010. Wind turbine efficiency optimization. Comparative study of controllers based on second order sliding modes. Int. J. Hydrogen Energy 35, 5934e5939. https://doi.org/10.1016/ j.ijhydene.2009.12.104. Frost, S.A., Balas, M.J., Wright, A.D., 2011. Mechatronics Generator speed regulation in the presence of structural modes through adaptive control using residual mode filters. Mechatronics 21, 660e667. https://doi.org/10.1016/ j.mechatronics.2011.01.006. Frost, S.A., Balas, M.J., Wright, A.D., 2009. Direct adaptive control of a utility-scale wind turbine for speed regulation. Int. J. Robust Nonlinear Control 19, 59e71. Geyler, M., Caselitz, P., 2007. Individual blade pitch control design for load reduction on large wind turbines. Eur. Wind Energy Conf. (EWEC 2007). Girbau-Llistuella, F., Sumper, A., Díaz-Gonz
alez, F., Galceran-Arellano, S., 2014. Flicker mitigation by reactive power control in wind farm with doubly fed induction generators. Int. J. Electr. Power Energy Syst. 55, 285e296. https:// doi.org/10.1016/j.ijepes.2013.09.016.
, G., Carranza, O., 2010. Maximum-power-point Gonz
alez, L.G., Figueres, E., Garcera tracking with reduced mechanical stress applied to wind-energy-conversionsystems. Appl. Energy 87, 2304e2312. https://doi.org/10.1016/ j.apenergy.2009.11.030. GWEC, 2016. Global wind report 2015. Gwec. Wind energy Technol. 75. Hafiz, F., Abdennour, A., 2015. Optimal use of kinetic energy for the inertial support from variable speed wind turbines. Renew. Energy 80, 629e643. https:// doi.org/10.1016/j.renene.2015.02.051. Hand, M.M., Balas, M.J., 2002. Systematic Controller Design Methodology for Variable-speed Wind Turbines. Hansen, M., Hansen, A., Larsen, T., 2005. Control Design for a Pitch-regulated, Variable Speed Wind Turbine. Riso National Laboratory. https://doi.org/ 10.1111/j.1523-1739.2010.01586.x. Harris, M., Hand, M., Wright, A.D., 2006. Lidar for Turbine Control. NREL Tech. Rep. NREL/TP-50, TP-500e39154. doi:NREL/TP-500-39154. Hau, E., von Renouard, H., 2005. Wind Turbines: Fundamentals, Technologies, Application, Economics. Springer Berlin Heidelberg. Henriksen, L.C., 2007. Model predictive control of a wind turbine. Math. Model 1e7. Henriksen, L.C., Hansen, M.H., Poulsen, N.K., 2012. Wind turbine control with constraint handling: a model predictive control approach. IET Control Theory Appl. 6, 1722. https://doi.org/10.1049/iet-cta.2011.0488. Hilloowala, R.M., Sharaf, a. M., 1994. Utility interactive wind energy conversion scheme with an asynchronous DC link using a supplementary control loop. IEEE Trans. Energy Convers. 9, 558e563. https://doi.org/10.1109/60.326477. Huang, P., Shawky, M., Moursi, E., Hasen, S.A., Member, S., 2015. Novel fault ridethrough scheme and control strategy for doubly fed induction generatorbased wind turbine. IEEE Trans. Energy Convers. 30, 635e645. Jain, A., Schildbach, G., Fagiano, L., Morari, M., 2015a. On the design and tuning of linear model predictive control for wind turbines. Renew. Energy 80, 664e673. https://doi.org/10.1016/j.renene.2015.02.057. Jain, B., Jain, S., Nema, R.K., 2015b. Control strategies of grid interfaced wind energy conversion system: an overview. Renew. Sustain. Energy Rev. 47, 983e996. https://doi.org/10.1016/j.rser.2015.03.063. Jelavi
c, M., Petrovi
c, V., Peric, N., 2010. Estimation based individual pitch control of wind turbine. Automatika 51, 181e192. Johnson, K., Wingerden, J.-W. Van, Balas, M.J., Molenaar, D.-P., 2011. Special issue on “Past, present and future modeling and control of wind turbines”. Mechatronics 21, 633. https://doi.org/10.1016/j.mechatronics.2011.03.005. Johnson, K.E., 2008. Adaptive Torque Control of Variable Speed Wind Turbines. doi: NREL/TP-500-36265. Jonkman, J., Butterfield, S., Musial, W., Scott, G., 2009. Definition of a 5-MW reference wind turbine for offshore system development. Contract 1e75. https:// doi.org/10.1002/ajmg.10175. Jukes, T.N., 2015. Smart control of a horizontal axis wind turbine using dielectric barrier discharge plasma actuators. Renew. Energy 80, 644e654. https:// doi.org/10.1016/j.renene.2015.02.047. Kassem, A.M., 2012. Robust voltage control of a stand alone wind energy conversion system based on functional model predictive approach. Int. J. Electr. Power Energy Syst. 41, 124e132. https://doi.org/10.1016/j.ijepes.2012.03.027. Kassem, A.M., Yousef, A.M., 2013. Voltage and frequency control of an autonomous hybrid generation system based on linear model predictive control. Sustain. Energy Technol. Assess. 4, 52e61. https://doi.org/10.1016/j.seta.2013.09.002. Kim, Y.-S., Chung, I.-Y., Moon, S.-I., 2015. Tuning of the PI controller parameters of a PMSG wind turbine to improve control performance under various wind speeds. Energies 8, 1406e1425. https://doi.org/10.3390/en8021406. Koerber, a, King, R., 2013. Combined feedback and feedforward control of wind turbines using state-constrained model predictive control. Control Syst.

E.J. Novaes Menezes et al. / Journal of Cleaner Production 174 (2018) 945e953 Technol. IEEE Trans. 21, 1117e1128. https://doi.org/10.1109/TCST.2013.2260749. Kota, S., 2008. Variable Geometry Airfoils and Active Flow Control. In: Laboratories, S.N. (Ed.), Proceedings of the IEA Topical Expert Meeting of Smart Structures for Large Wind Turbine Rotor Blades. Albuquerque, USA. Kragh, K., Hansen, M., 2010. Model predictive individual pitch control based on local inflow measurements. 7th PhD Semin. Wind Energy Eur. 3e6. Kumar, A., Stol, K., 2009. Scheduled model predictive control of a wind turbine. In: 47th AIAA Aerospace Sciences Meeting Including the New Horizons Forum and Aerospace Exposition, Aerospace Sciences Meetings. American Institute of Aeronautics and Astronautics. https://doi.org/10.2514/6.2009-481. Kusiak, A., Li, W., Song, Z., 2010. Dynamic control of wind turbines. Renew. Energy 35, 456e463. https://doi.org/10.1016/j.renene.2009.05.022. Laks, J., Pao, L., Wright, A., Kelley, N., Jonkman, B., 2011. The use of preview wind measurements for blade pitch control. Mechatronics 21, 668e681. https:// doi.org/10.1016/j.mechatronics.2011.02.003. Laks, J.H., Pao, L.Y., Wright, A.D., 2009. Control of wind Turbines: past, present, and future. Am. Control Conf. 2096e2103. doi:10.1109/ACC.2009.5160590. Li, H., Steurer, M., Shi, K.L., Woodruff, S., Zhang, D., 2006. Development of a unified design, test, and research platform for wind energy systems based on hardware-in-the-loop real-time simulation. IEEE Trans. Ind. Electron 53, 1144e1151. https://doi.org/10.1109/TIE.2006.878319. Li, P., Song, Y., Li, D., Cai, W., Zhang, K., Review, A., Power, G.W., 2015. Control and monitoring for grid-friendly wind turbines. Res. Overv. Suggest. Approach 30, 1979e1986. Lubosny, Z., Lubosny, Z., Bialek, J.W., Bialek, J.W., 2007. Supervisory control of a wind farm. IEEE Trans. Power Syst. 22, 985e994. Luzar, M., Witczak, M., 2014. Robust MPC for a non-linear system e a neural network approach. J. Phys. Conf. Ser. 570, 32002. https://doi.org/10.1088/17426596/570/3/032002. Manonmani, N., Kausalyadevi, P., 2014. A review of maximum power extraction techniques for wind energy conversion systems. Int. J. Innov. Sci. Eng. Technol. 1, 597e604. Manwell, J.F., McGowan, J.G., Rogers, A.L., 2010. Wind Energy Explained: Theory, Design and Application. Wiley. Marschner, V., Michael, J., Liersch, J., 2014. A process for providing positive primary control power by wind turbines. J. Phys. Conf. Ser. 570, 52002. https://doi.org/ 10.1088/1742-6596/570/5/052002. Mazenc, F., Queiroz, M.S. De, Malisoff, M., Gao, F., 2015. Robust MPC tower damping for variable speed wind turbines. IEEE Trans. Control Syst. Technol. 23, 290e296. Miller, L., 1995. Experimental Investigation of Aerodynamic Devices for Wind Turbine Rotational Speed Control. NREL Technical Report, 1995. Miller, N.W., Delmerico, R.W., Kuruvilla, K., Shao, M., 2012. Frequency responsive controls for wind plants in grids with wind high penetration. In: IEEE Power Energy Soc. Gen. Meet., pp. 1e7. https://doi.org/10.1109/PESGM.2012.6344994. Mohamed, T.H., Morel, J., Bevrani, H., Hiyama, T., 2012. Model predictive based load frequency control_design concerning wind turbines. Int. J. Electr. Power Energy Syst. 43, 859e867. https://doi.org/10.1016/j.ijepes.2012.06.032. Muljadi, E., 2001. Pitch-controlled variable-speed wind turbine generation. IEEE Trans. Ind. Appl. 37, 240e246. https://doi.org/10.1109/28.903156. Munteanu, I., Bacha, S., Bratcu, A.I., Roye, D., 2008. Energy-reliability optimization of wind energy conversion systems by sliding mode control. IEEE Trans. Energy Convers. 23, 975e985. Nakafuji, D.T.Y., van Dam, C.P., Smith, R.L., Collins, S.D., 2001. Active load control for airfoils using microtabs. J. Sol. Energy Eng. 123, 282e289. Namik, H., Stol, K., 2011. Performance analysis of individual blade pitch control of offshore wind turbines on two floating platforms. Mechatronics 21, 691e703. https://doi.org/10.1016/j.mechatronics.2010.12.003. Navalkar, S.T., van Wingerden, J.W., van Solingen, E., Oomen, T., Pasterkamp, E., van Kuik, G. a M., 2014. Subspace predictive repetitive control to mitigate periodic loads on large scale wind turbines. Mechatronics 24, 916e925. https://doi.org/ 10.1016/j.mechatronics.2014.01.005. € ffker, D., 2016. State-of-the-art in wind turbine control: trends and Njiri, J.G., So challenges. Renew. Sustain. Energy Rev. 60, 377e393. https://doi.org/10.1016/ j.rser.2016.01.110. Oudah, A.I., Mohd, I., Hameed, a, 2014. Wind turbines control: features and trends. Mod. Appl. Sci. 8, 272e295. https://doi.org/10.5539/mas.v8n6p272. Petrovi
c, V., Jelavi
c, M., Baoti
c, M., 2015. Advanced control algorithms for reduction of wind turbine structural loads. Renew. Energy 76, 418e431. https://doi.org/ 10.1016/j.renene.2014.11.051. Plumley, C.E., Graham, M., Leithead, W.E., Bossanyi, E., Jamieson, P., 2014. Supplementing wind turbine pitch control with a trailing edge flap smart rotor. In: 3rd Renew. Power Gener. Conf. (RPG 2014). https://doi.org/10.1049/cp.2014.0919, 8.34e8.34. Poultangari, I., Shahnazi, R., Sheikhan, M., 2012. RBF neural network based PI pitch controller for a class of 5-MW wind turbines using particle swarm optimization algorithm. ISA Trans. 51, 641e648. https://doi.org/10.1016/j.isatra.2012.06.001.

953

Pudjianto, D., Pudjianto, D., Ramsay, C., Ramsay, C., Strbac, G., Strbac, G., 2007. Virtual power plant and system integration of distributed energy resources. Renew. Power Gener. IET 1, 10e16. https://doi.org/10.1049/iet-rpg. Quandt, G., Migliore, P., 1996. Wind Turbine Trailing-edge Aerodynamic Brake Design. Rajendran, S., Jena, D., 2014. Control of variable speed variable pitch wind turbine at above and below rated wind speed. J. Wind Energy. https://doi.org/10.1155/ 2014/709128. Raza Kazmi, S.M., Goto, H., Guo, H.-J., Ichinokura, O., 2011. A novel algorithm for fast and efficient speed-sensorless maximum power point tracking in wind energy conversion systems. Ind. Electron. IEEE Trans. 58, 29e36. https://doi.org/ 10.1109/TIE.2010.2044732. Rodriguez-Amenedo, J.L., Arnalte, S., Burgos, J.C., 2002. Automatic generation control of a wind farm with variable speed. Energy Conversion. IEEE Trans. 17, 279e284. https://doi.org/10.1109/TEC.2002.1009481. Rohatgi, J., Vaughn, N., 1994. Wind Characteristics: an Analysis for the Generation of Wind Power. Alternative Energy Institute. Saqib, M. a, Saleem, A.Z., 2015. Power-quality issues and the need for reactivepower compensation in the grid integration of wind power. Renew. Sustain. Energy Rev. 43, 51e64. https://doi.org/10.1016/j.rser.2014.11.035. Saravanakumar, R., Jena, D., 2015. Validation of an integral sliding mode control for optimal control of a three blade variable speed variable pitch wind turbine. Int. J. Electr. Power Energy Syst. 69, 421e429. https://doi.org/10.1016/ j.ijepes.2015.01.031. Schlipf, D., Fleming, P., Haizmann, F., Scholbrock, A., Hofs€ aß, M., Wright, A., Cheng, P.W., 2014. Field testing of feedforward collective pitch control on the CART2 using a nacelle-based lidar scanner. J. Phys. Conf. Ser. 555, 12090. https:// doi.org/10.1088/1742-6596/555/1/012090. Shan, M., Jacobsen, J., Adelt, S., 2013. Field Testing and Practical Aspects of Load Reducing Pitch Control Systems for a 5 MW Offshore Wind Turbine. Ewea 2013. Sloth, C., Esbensen, T., Stoustrup, J., 2011. Robust and fault-tolerant linear parameter-varying control of wind turbines. Mechatronics 21, 645e659. https://doi.org/10.1016/j.mechatronics.2011.02.001. Smit, J., Bernhammer, L.O., Navalkar, S.T., Bergami, L., Gaunaa, M., 2015. Sizing and control of trailing edge flaps on a smart rotor for maximum power generation in low fatigue wind regimes. Wind Energy. https://doi.org/10.1002/we.1853. Spencer, M.D., Stol, K.A., Unsworth, C.P., Cater, J.E., Norris, S.E., 2013. Model predictive control of a wind turbine using short-term wind field predictions. Wind Energy 16, 417e434. https://doi.org/10.1002/we.1501. van Solingen, E., Fleming, P.A., Scholbrock, A., van Wingerden, J.W., 2015. Field testing of linear individual pitch control on the two-bladed controls advanced research turbine. Wind Energy. https://doi.org/10.1002/we.1841. Vogler-Finck, P.J.C., Früh, W.-G., 2015. Evolution of primary frequency control requirements in Great Britain with increasing wind generation. Int. J. Electr. Power Energy Syst. 73, 377e388. https://doi.org/10.1016/j.ijepes.2015.04.012. Vries, H. De, Weide, E.T. a Van Der, Hoeijmakers, H.W.M., 2014. Synthetic jet actuation for load control. J. Phys. Conf. Ser. 555, 12026. https://doi.org/10.1088/ 1742-6596/555/1/012026. Wang, L., 2009. Model Predictive Control System Design and Implementation Using MATLAB®, Advances in Industrial Control. Springer. Wang, N., Johnson, K.E., Wright, A.D., 2013. Comparison of strategies for enhancing energy capture and reducing loads using LIDAR and feedforward control. IEEE Trans. Control Syst. Technol. 21, 1129e1142. Wang, N., Johnson, K.E., Wright, A.D., Carcangiu, C.E., 2014. Lidar-assisted wind turbine feedforward torque controller design below rated. Proc. Am. Control Conf. 3728e3733. Wang, S., Seiler, P.J., 2014. Gain scheduled active power control for wind turbines. In: 32nd ASME Wind Energy Symposium, AIAA SciTech. American Institute of Aeronautics and Astronautics. https://doi.org/10.2514/6.2014-1220. Yao, X., Guan, L., Guo, Q., Ma, X., 2010. RBF neural network based self-tuning PID pitch control strategy for wind power generation system. In: Computer, Mechatronics, Control and Electric Engineering, pp. 482e485. Yaoqin, J.Y.J., Zhongqing, Y.Z.Y., Binggang, C.B.C., 2002. A new maximum power point tracking control scheme for wind generation. Proceedings. Int. Conf. Power Syst. Technol. 1, 144e148. https://doi.org/10.1109/ICPST.2002.1053521. Yingcheng, X., Nengling, T., 2012. System frequency regulation in doubly fed induction generators. Int. J. Electr. Power Energy Syst. 43, 977e983. https:// doi.org/10.1016/j.ijepes.2012.05.039. Yingcheng, X., Nengling, T., 2011. Review of contribution to frequency control through variable speed wind turbine. Renew. Energy 36, 1671e1677. https:// doi.org/10.1016/j.renene.2010.11.009. Yu Zou, Elbuluk, M.E., Sozer, Y., 2013. Stability analysis of maximum power point tracking ( MPPT ) method in wind power systems. IEEE Trans. Ind. Appl. 49, 1129e1136. Zhu, Y., Cheng, M., Hua, W., Wang, W., 2012. A novel maximum power point tracking control for permanent magnet direct drive wind energy conversion systems. Energies 5, 1398e1412. https://doi.org/10.3390/en5051398.