Search

Wind farm flow control (WFFC) is a topic of interest at several research institutes, industry and certification agencies world-wide. For reliable performance assessment of the technology, the efficiency and the capability of the models applied to WFFC should be carefully evaluated. To address that, FarmConners consortium has launched a common benchmark for code comparison under controlled operation to demonstrate its potential benefits such as increased power production. The benchmark builds on available data sets from previous field campaigns, wind tunnel experiments and high-fidelity simulations. Within that database, 4 blind tests are defined and 13 participants in total have submitted results for the analysis of single and multiple wake under WFFC. Some participants took part in several blind tests and some participants have implemented several models. The observations and/or the model outcomes are evaluated via direct power comparisons at the upstream and downstream turbine(s), as well as the power gain at the wind farm level under wake steering control strategy. Additionally, wake loss reduction is also analysed to support the power performance comparison, where relevant. Majority of the participating models show good agreement with the observations or the reference high-fidelity simulations, especially for lower degrees of upstream misalignment and narrow wake sector. However, the benchmark clearly highlights the importance of the calibration procedure for control-oriented models. The potential effects of limited controlled operation data in calibration is particularly visible via frequent model mismatch for highly deflected wakes, as well as the power loss at the controlled turbine(s). In addition to the flow modelling, sensitivity of the predicted WFFC benefits to the turbine representation and the implementation of the controller is also underlined. FarmConners benchmark is the first of its kind to bring a wide variety of data sets, control settings and model complexities for the (initial) assessment of farm flow control benefits. It forms an important basis for more detailed benchmarks in the future with extended control objectives to assess the true value of WFFC.

The turbulent conditions a wind energy converter is working in are resulting in strong angle of attack (AoA) fluctuations on the blades of the wind turbine rotor. As a consequence the lift force is constantly varying and phenomena as dynamic stall are occurring, inducing high turbine loads. In this study we investigate this phenomenon by means of particle image velocimetry (PIV) measurements on a rotating model wind turbine. The experiments took place in the 3 m times 3 m wind tunnel of the University of Oldenburg utilizing the Model Wind Turbine of Oldenburg (MoWiTO) [1] and a high-speed stereo PIV system. Figure 1 (a) is showing a picture of the used setup including an active grid for turbulence generation [2]. The flow around the three airfoil sections was captured at the 12 o’clock position using cyclic triggering at a sampling frequency of 5 kHz while the turbine was operating at a rotating speed of 480 rpm. Different positions at the blade root at 25%, 30% and 35% of the span were examined this way. To induce dynamic stall the blades of the turbine were pitched in a sinusoidal movement between -1° to 3° using the turbine’s individual pitch control. Stall clearly occurs at a pitch angle of -1°, corresponding to AoA of 13°, 10° and 9° for the investigated radial positions at steady operation. By shifting the phase of the pitching movement in respect to one rotation of the rotor different AoA can be visualized at the 12 o’clock position of the turbine blades. In Figure 1 (b) a result of these measurements is shown. The inflow is coming from the right and the turbine blade moving from the bottom to the top. Colour coded we can see the vorticity of the flow field around the wind turbine blade. Clearly we observe a separation of the flow at the suction side of the airfoil and the subsequent wake which is advected by the main flow. With this novel setup dynamic stall can be induced and examined thoroughly using a high frequent and non-intrusive measurement method. This results in a better understanding of the flow around the blade root of a down scaled model wind turbine in controlled wind tunnel conditions. [1] Berger et al, Journal of Physics Conference Series. 1104, 012026 (2018). [2] Kröger et al., Journal of Physics Conference Series. 1037, 052002 (2018).

The current trend toward larger wind turbine rotors leads to high periodic loads across the components due to the non-uniformity of inflow across the rotor. To address this, we introduce a blade-mounted lidar on each blade to provide a preview of inflow wind speed that can be used as a feedforward control input for the mitigation of such periodic blade loads. We present a method to easily determine blade-mounted lidar parameters, such as focus distance, telescope position, and orientation on the blade. However, such a method is accompanied by uncertainties in the inflow wind speed measurement, which may also be due to the induction zone, wind evolution, “cyclops dilemma”, unidentified misalignment in the telescope orientation, and the blade segment orientation sensor. Identification of these uncertainties allows their inclusion in the feedback–feedforward controller development for load mitigation. We perform large-eddy simulations, in which we simulate the blade-mounted lidar including the dynamic behaviour and the induction zone of one reference wind turbine for one above-rated inflow wind speed. Our calculation approach provides a good trade-off between a fast and simple determination of the telescope parameters and an accurate inflow wind speed measurement. We identify and model the uncertainties, which can then be directly included in the feedback–feedforward controller design and analysis. The rotor induction effect increases the preview time, which needs to be considered in the controller development and implementation.

To enable further increases in wind turbine dimensions, and thus to lower the levelized cost of energy, active and passive concepts for load reduction have to be further developed. One of such concepts is individual pitch control, which has been thoroughly analysed and validated in simulations and experiments (e.g., see [1] and references therein), and is currently used by some modern multi-megawatt wind turbines. On the other hand, although well known from aeronautics, concepts based on rotor blades with trailing edge flaps or leading edge slats have still not found any commercial application in wind energy. One obstacle is seen in the lack of a thorough and convincing validation of such concepts. In this work, we present our wind tunnel setup for experimental validation of advanced concepts for wind turbine load reduction, and analyse the interaction between rigid slats and the wind turbine control system in different inflow conditions. The wind tunnel at the University of Oldenburg has a nozzle of 3m by 3m and an active grid, which can impress a wide range of tailored turbulent flow conditions and coherent gusts on the wind flow [2]. The fully controllable and aerodynamically scaled version of a 5MW reference turbine MoWiTO (Model Wind Turbine Oldenburg) with a rotor diameter of 1.8m features the measurement of flapwise blade root and tower bending moments, rotor speed, and torque[3]. MoWiTO can be operated with two sets of blades – a set with rigid leading edge slats[4], and a reference set without slats. Additionally, the real-time hardware enables implementation of different control algorithms, but for the purpose of this work, two control algorithms will be used: a baseline controller for torque and collective pitch control, and an individual pitch controller designed to reduce once-per-revolution blade bending moments. Using the active grid, the capability of the rigid slats in combination with individual pitch control will be analysed in different inflow conditions, including turbulent and sheared flows, and wind gusts. [1] Bossanyi et al., IEEE Trans. on Control System Technology 21, 1067 (2013) [2] Kröger et al., Journal of Physics: Conference Series 1037, 052002 (2018) [3] Berger et al., Journal of Physics: Conference Series 1104, 012026 (2018) [4] Manso Jaume et al., Deutsche Windenergie Konferenz (2015)

In this paper, we extend our closed-loop optimal control framework for wind farms to minimize wake-induced power losses. We develop an adjoint-based model predictive controller which employs a medium-fidelity 2D dynamic wind farm model. The wind turbine axial induction factors are considered here as the control inputs to influence the overall performance by taking wake interactions of the wind turbines into account. A constrained optimization problem is formulated to maximize the total power production of a given wind farm. An adjoint approach as an efficient tool is utilized to compute the gradient for such a large-scale system. The computed gradient is then modified to deal with the defined final set and practical constraints on the wind turbine control inputs. The performance of the wind farm controller is examined for a more realistic test case, a layout of a 2 x 3 wind farm with dynamical changes in wind direction. The effectiveness of the proposed approach is studied through simulations.

In this thesis, control algorithms for reduction of wind turbine structural loads in strong winds have been researched. Such reduction is essential for prolongation of expected lifetime of wind turbine and for enabling further increase of its dimensions and rated power. Two approaches are considered for achieving reduction of wind turbine structural loads. Firstly, a control algorithm is designed for very high wind speeds (typically present during storms). This control algorithm enables wind turbine operation even when standard control system has to shut the wind turbine down thus improving wind farm behaviour in the electrical grid. An online optimisation for determining worst-case prediction of the wind speed is used, and based on such prediction, wind turbine power reference is changed to avoid excessive wind turbine structural loads. A second approach for reduction of wind turbine structural loads described in the thesis is focused on wind speeds above nominal wind speed. Standard individual pitch control algorithms and higher harmonic control algorithms are extended to enable the reduction of structural loads caused by rotor asymmetry. Furthermore, appropriate transformations of structural loads are defined that enable the use of structural load measurements from any part of the wind turbine, both rotating and non-rotating, to achieve reduction of structural loads. The reduction of structural loads is incorporated in a multi-criteria controller to improve wind turbine performance, with special emphasis on the satisfaction of pitch actuator constraints. To this end, a procedure is derived that allows better integration of the pitch actuator limitations (when considering structural loads reduction) within the model predictive control strategy. Finally, two scaled wind turbine models suitable for experimental validation of control algorithms in wind tunnels are presented. Such scaled models bridge the gap between numerical simulations and field tests, thus improving the procedure for validation of new control algorithms. Reported experimental results indicate that the presented scaled wind turbine models can be used for this purpose. U ovoj disertaciji, istraženo je smanjenje strukturnih opterećenja vjetroagregata korištenjem naprednih algoritama upravljanja. Smanjenje opterećenja je nužno za produljenje životnog vijeka vjetroagregata, odnosno za omogućavanje daljnjeg porasta dimenzija i nazivne snage vjetroagregata. Razmatrana su dva pristupa za smanjivanje strukturnih opterećenja vjetroagregata. Prvi pristup odnosi se na algoritam upravljanja projektiran za iznimno jake vjetrove (za olujno vrijeme). Razvijeni algoritam upravljanja omogućuje rad vjetroagregata pri brzinama vjetra kod kojih standardni algoritmi upravljanja isključuju vjetroagregat, čime se postiže povoljnije vladanje vjetroelektrane u elektroenergetskom sustavu. Najprije se koristi linearni optimizacijski problem za predviđanje vjetra koji bi uzrokovao najveća opterećenja, a potom se prilagođava referenca snage vjetroagregata kako bi se izbjegla prekomjerna opterećenja i u slučaju da se ostvari predviđeni vjetar. Drugi pristup za smanjenje strukturnih opterećenja vjetroagregata fokusiran je na brzine vjetra iznad nazivnog iznosa. Standardni algoritmi za pojedinačno zakretanje lopatica su prošireni kako bi se omogućilo smanjenje opterećenja uzrokovanih asimetrijama rotora. Nadalje, definirane su prikladne transformacije opterećenja koje omogućuju smanjenje strukturnih opterećenja korištenjem mjerenja opterećenja u proizvoljnim točkama vjetroagregata. Smanjenje strukturnih opterećenja je uključeno u višekriterijski regulator s ciljem postizanja kvalitetnijeg vladanja vjetroagregata, pri čemu je poseban naglasak dan na poštivanje ograničenja aktuatora za zakret lopatica. U tu svrhu, razvijen je postupak koji omogućava integraciju ograničenja aktuatora za zakret lopatica (uz korištenje pojedinačnog zakretanja lopatica) u algoritme modelsko prediktivnog upravljanja. Naposljetku, opisana su dva skalirana modela vjetroagregata prikladna za eksperimentalnu provjeru algoritama upravljanja u zračnim tunelima. Takvi skalirani modeli predstavljaju poveznicu između numeričkih simulacija i eksperimenata na terenu te unaprjeđuju postupak provjere novih algoritama upravljanja. Prikazani eksperimentalni rezultati pokazuju da su opisani skalirani modeli vjetroagregata prikladni za tu svrhu.

U studiji je obrađena mogućnost predikcije satne potrošnje električne energije po distribucijskim područjima primjenom neuronskih mreža. Provedena je analiza relevantnosti ulaznih varijabli metodom parcijalne zajedničke informacije te su analizirane različite strukture neuronskih mreža. Studija se prvenstveno bavi kratkoročnim prediđanjem potrošnje električne energije, a analizirana je i mogućnost srednjoročne predikcije.

Modern wind turbines operate in a wide range of wind speeds. Power contained in the wind is proportional to the third power of wind speed and therefore increases rapidly with increase of wind speed. To enable wind turbine operation in such a variety of operating conditions, sophisticated control and estimation algorithms are needed. In this paper, a method for wind turbine state and parameter estimation is proposed. The described estimation is experimentally tested on laboratory wind turbine.

U radu je opisana primjena LabVIEW programskog alata za razvoj Hardware-in-the-Loop strukture upravljanja brzinom vrtnje vjetroagregata. Klasični algoritmi upravljanja vjetroagregatom zasnivaju se na upravljanju brzinom vrtnje vjetroagregata. Pritom se razlikuju dva različita radna područja u kojima se vjetroagregat može nalaziti: 1) područje slabih vjetrova u kojem je važno maksimalno iskoristiti raspoloživu energiju vjetra ; u ovom području rada lopatice vjetroagregata miruju i postavljene su na kut koji osigurava optimalnu pretvorbu energije, dok se elektromehanički moment generatora može postavljati prema preglednoj tablici u ovisnosti o brzini vrtnje ; i 2) područje jakih vjetrova u kojem je snaga vjetra prevelika i potrebno je ograničiti energiju koja se preuzima iz vjetra. Razvijena struktura upravljanja provjerena je na vjetroagregatu koji je dio Laboratorija za obnovljive izvore energije na Fakultetu elektrotehnike i računarstva Sveučilišta u Zagrebu.

U izvješću je opisan sustav upravljanja laboratorijskim vjetroagregatom u Laboratoriju za obnovljive izvore energije (LARES) na FER-u. Za upravljanje je korištena LabVIEW platforma od National Instrumentsa.

U radu je opisano modeliranje vjetroagregata megavatne klase u Matlab/Simulink okruženju. Validacija modela provedena je uz pomoć profesionalnog simulacijskog alata GH Bladed

U ovom radu opisane su metode upravljanja vjetroagregatom s promjenjivom brzinom vrtnje i regulacijom zakretanjem lopatica. Najprije je dan pregled fizikalnih osnova rada vjetroagregata i modela korištenog pri projektiranju sustava upravljanja. Zatim su opisani klasični algoritmi upravljanja vjetroagregatom u dva različita radna područja - ispod i iznad nazivne brzine vjetra. Razmotrena je i unaprijedna kompenzacija promjene brzine vjetra korištenjem estimatora brzine vjetra te upravljački algoritmi za smanjenje njihanja tornja. Kod opisa pojedinih upravljačkih algoritama, najprije je dan načelni opis, a potom su razmotreni neki praktični aspekti implementacije algoritama.