A data-driven model for wind plant power optimization by yaw control

Gebraad, Pieter; Teeuwisse, F. W.; Wingerden, Jan‐Willem van; Fleming, Paul; Ruben, Shalom; Marden, Jason R.; Pao, Lucy Y.

doi:10.1109/acc.2014.6859118

Cited by 80 publications

(70 citation statements)

References 15 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…We will assume a 3 parameter profile. It has been observed that a turbine wake can be divided into multiple regions 1,2,6,7,13,14 , each with their own energy diffusion processes. For convenience we will consider two distinct regions; A core flow where the velocity deficit is constant, and a shear disk where velocity recovers to the free stream value at the wake edge as in figure- Figure (xx7?).…”

Section: Parameter Fitting Based On Integral Quantitiesmentioning

confidence: 99%

“…This method of control has been shown to be more effective than TSR control at increasing cluster efficiency and reducing unsteady loads [XX]. 13 An analytical model of yaw deflection has been derived by Jimenez 15 based on high fidelity CFD solutions. The model inputs are the non-yawed axial induction factor aj and the turbine yaw angle γj.…”

Section: Related Workmentioning

confidence: 99%

See 1 more Smart Citation

A Fast Differential Deficit Control Volume Approach for Modeling Turbine-Turbine Interactions

DiPaola

Willis

2016

46th AIAA Fluid Dynamics Conference

View full text Add to dashboard Cite

Several computational tools exist for wind turbine design and wind farm planning/layout ranging from time-intensive computational fluid dynamics (CFD) to lower-fidelity velocity deficit wake models. Analytical and reduced order models (ROMs) are computationally inexpensive but lack the necessary physics to make trend accurate, time-dependent predictions; conversely, high fidelity, Large Eddy Simulations (LES) CFD remains prohibitively expensive and complex for routine wind farm design, optimization and real-time control. This paper presents a one-dimensional, axially symmetric reduced order model that uses integral conservation laws coupled with empirical closure relations for wind turbine wake evolution. I. IntroductionThe 2008 assessment titled 20% Wind Energy by 2030; Increasing Wind Energy's Contribution to U.S. Electricity Supply 8 , investigates the feasibility of a 20% wind generation scenario, which would require mobilization and growth in areas of turbine technology, manufacturing and raw materials, transmission and integration, siting, environmental impacts, and markets. The report was a collaboration of the U.S. Department of Energy and various laboratories including the National Renewable Energy Laboratory and Sandia National Laboratory among others, as well as engineering and consulting firms and the American Wind Energy Association (AWEA) made up of wind manufacturers and suppliers, developers, and electric utilities. The collaborative highlights the theoretical need for immense growth in installed capacity, from 11.6 GW in 2006 to 305 GW in 2030, but concludes the scenario economically viable and identifies initiatives to improve wind turbine performance. These initiatives are aimed at avoiding problems before installation by improving component reliability, full-scale testing methods prior to commercial deployment, validation of design tools and development of standardized design criteria and specifications. More robust methods of performance assurance that can measure, verify, and evaluate individual turbine and largescale wind-plant performance will become increasingly important.

show abstract

Section: Parameter Fitting Based On Integral Quantitiesmentioning

confidence: 99%

Section: Related Workmentioning

confidence: 99%

A Fast Differential Deficit Control Volume Approach for Modeling Turbine-Turbine Interactions

DiPaola

Willis

2016

46th AIAA Fluid Dynamics Conference

View full text Add to dashboard Cite

show abstract

“…In the above power expression, we have the factor cos(γ n ) 3 with exponent 3. In literature such as, e.g., (Gebraad et al, 2014a) and (Medici, 2005) (page 37), numerical values for the exponent were given according to LES and wind tunnel data, respectively. However, to date, the exact value for it is still under research and since this is outside the scope of this study, the value of the exponent will be three.…”

Section: Powermentioning

confidence: 99%

A control-oriented dynamic wind farm model: WFSim

Boersma¹,

Doekemeijer²,

Vali³

et al. 2017

Preprint

View full text Add to dashboard Cite

Abstract.Wind turbines are often sited together in wind farms as it is economically advantageous. Controlling the flow within wind farms to reduce the fatigue loads and provide grid facilities such as the delivery of a demanded power is a challenging control problem due to the underlying time-varying nonlinear wake dynamics. It is therefore important to use the closed-loop control paradigm since it can partially account for model uncertainty and, in addition, it can deal with unknown disturbances. State-5 of-the-art closed-loop dynamic wind farm controllers are based on computationally expensive wind farm models, which make these methods suitable for analysis though unsuitable for online control. The latter is important, because it allows for model adaptation to the time-varying atmospheric conditions using SCADA measurements. As a consequence, more reliable control settings can be evaluated.In this paper, a dynamic wind farm model suitable for online wind farm control will be presented. The derivation of the 10 control-oriented dynamic wind farm model starts with the three-dimensional Navier-Stokes equations. Then, terms involving the vertical dimension will be estimated in order to partially compensate for neglecting the vertical dimension or neglected such that a 2D-like dynamic wind farm model will be obtained. Sparsity of and structure in the system matrices make this model relatively computational inexpensive hence suitable for online closed-loop controller synthesis including model parameter updates. Flow and power data evaluated with the wind farm model presented in this work will be validated with high fidelity 15 flow data.

show abstract

“…In [4] a simple yet widely used wake model is employed to model the wake effect, while [3] uses a finite volume discretization of the Navier-Stokes equation for the wind flow model. In another method which is presented in [7], the authors suggest deflecting wakes by yaw-misalignment in order to reduce their effect on the downstream wind turbines.…”

Section: Introductionmentioning

confidence: 99%

Turbine Control strategies for wind farm power optimization

Mirzaei

Göçmen

Giebel

et al. 2015

2015 American Control Conference (ACC)

View full text Add to dashboard Cite

In recent decades there has been increasing interest in green energies, of which wind energy is the most important one. In order to improve the competitiveness of the wind power plants, there are ongoing researches to decrease cost per energy unit and increase the efficiency of wind turbines and wind farms. One way of achieving these goals is to optimize the power generated by a wind farm. One optimization method is to choose appropriate operating points for the individual wind turbines in the farm. We have made three models of a wind farm based on three difference control strategies. Basically, the control strategies determine the steady state operating points of the wind turbines. Except the control strategies of the individual wind turbines, the wind farm models are similar. Each model consists of a row of 5MW reference wind turbines. In the models we are able to optimize the generated power by changing the power reference of the individual wind turbines. We use the optimization setup to compare power production of the wind farm models. This paper shows that for the most frequent wind velocities (below and around the rated values), the generated powers of the wind farms are different. This means that choosing an appropriate control strategy for the individual wind turbines will result in an increased power production of the wind farm.

show abstract

A data-driven model for wind plant power optimization by yaw control

Cited by 80 publications

References 15 publications

A Fast Differential Deficit Control Volume Approach for Modeling Turbine-Turbine Interactions

A Fast Differential Deficit Control Volume Approach for Modeling Turbine-Turbine Interactions

A control-oriented dynamic wind farm model: WFSim

Turbine Control strategies for wind farm power optimization

Contact Info

Product

Resources

About