OPERATION OF A WIND TURBINE USING OPTIMIZED PARAMETERS

Abstract

Provided is a method for controlling a wind turbine, in particular an electric generator of said wind turbine. The method includes an optimization during which a suitable operating parameter for controlling said wind turbine or generator thereof is determined, in particular in an iterative manner. The optimization includes providing a multidimensional space comprising a plurality of parameters; providing an objective function for said multidimensional space, e.g., a simplex has a shape of a triangle or a tetrahedron; and determining one parameter of said multidimensional space as a suitable operating parameter by applying said objective function to said multidimensional space, in particular in an iterative manner. The method includes selecting a suitable operating parameter as an operating parameter for said wind turbine or generator thereof; and operating said wind turbine or generator based on said operating parameter, in particular by controlling a converter connected to said generator.

Claims

1. A method for controlling a wind turbine, comprising: performing an optimization to determine a first operating parameter for controlling the wind turbine, performing the optimization including: providing a multidimensional space comprising a plurality of parameters; providing an objective function for the multidimensional space; and determining one parameter of the multidimensional space as the first operating parameter by iteratively applying the objective function to the multidimensional space; selecting the first operating parameter as an operating parameter for the wind turbine; and operating the wind turbine, based on the operating parameter by controlling a converter connected to an electric generator of the wind turbine.

2. The method for controlling the wind turbine according to claim 1, wherein controlling the wind turbine includes controlling the electric generator of the wind turbine.

3. The method for controlling the wind turbine according to claim 1, wherein the objective function for the multidimensional space is a simplex that is triangle-shaped or tetrahedron-shaped.

4. The method for controlling the wind turbine according to claim 1, comprising: performing the optimization while the electric generator is rotating and/or generating electrical power.

5. The method for controlling the wind turbine according to claim 1, wherein the optimization is a mathematical optimization and/or numerical optimization.

6. The method for controlling the wind turbine according to claim 1, wherein the multidimensional space is provided as a circular grid having equal sized cells.

7. The method for controlling the wind turbine according to claim 6, wherein the equal sized cells have a triangular shape.

8. The method for controlling the wind turbine according to claim 1, comprising: determining the one parameter of the multidimensional space using a direct search method or an Nelder-Mead method.

9. The method for controlling the wind turbine according to claim 1, wherein the operating parameter is a minimum or a maximum of the objective function.

10. The method for controlling the wind turbine according to claim 1, wherein the optimization has at least a first mode and a second mode, and the first and second modes each having each having a different number of points that are measured.

11. The method for controlling the wind turbine according to claim 10, comprising: selecting the first mode or the second mode depending on a power ramp.

12. The method for controlling the wind turbine according to claim 1, comprising: determining whether the first operating parameter results in an operating current above a threshold; and in response to the first operating parameter resulting in the operating current above the threshold, refraining from selecting the first operating parameter as the operating parameter for the wind turbine.

13. The method for controlling the wind turbine according to claim 1, wherein performing the optimization includes: storing measurement points in a look-up table.

14. A wind turbine, comprising: the electric generator that is controlled according to the method as claimed in claim 1.

15. The wind turbine according to claim 14, wherein the electric generator is a permanent magnet synchronous having an active rectifier, and wherein said active rectifier is configured to control a current of the electric generator.

16. The wind turbine according to claim 15, wherein the active rectifier is configured to control an amplitude and/or a phase angle of the current of the electric generator.

17. The wind turbine according to claim 14, comprising: a wind turbine controller configured to store measurement points.

18. The wind turbine according to claim 17, comprising: a sensor positioned on a hub of the wind turbine.

19. The wind turbine according to claim 18, wherein the sensor is configured to detect a vibration or a sound of the wind turbine, the electric generator and/or a wind turbine blade.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0061] With reference to the appended drawings, below follow a more detailed description of embodiments of the invention cited as examples.

[0062] In the drawings:

[0063] FIG. 1 shows a wind turbine in an embodiment,

[0064] FIG. 2 shows another view of said wind turbine,

[0065] FIG. 3 shows a flowchart of the proposed method,

[0066] FIG. 4A illustrates a multidimensional space,

[0067] FIG. 4B illustrates a multidimensional space during optimization,

[0068] FIG. 5 shows a flow chart of the optimization,

[0069] FIG. 6 shows a look-up table.

DETAILED DESCRIPTION

[0070] FIG. 1 shows a wind turbine 100.

[0071] Said wind turbine 100 comprises a tower 102 and a nacelle 104. Arranged on said nacelle 104 is an aerodynamic rotor 106 having a hub 107 with three rotor blades 108 and a spinner 110.

[0072] During operation, said aerodynamic rotor 106 is set in a rotating motion by the wind and thereby driving a generator in said nacelle 104.

[0073] Said generator is preferably a synchronous generator with an active rectifier, designed to control a current of said generator.

[0074] Said wind turbine 100 also comprises a sensor 200, in particular an accelerometer, placed in said hub 107, in particular on an inner side of said hub 107 next to a pitch bearing or a stator ring of the generator.

[0075] Said sensor 200 is designed for detecting a vibration or a sound of said wind turbine 100 and/or said electric generator and/or a wind turbine blade 108.

[0076] FIG. 2 shows another view of said wind turbine 100, in particular as shown in FIG. 1.

[0077] Said rotor blades 108 are mechanically connected to said generator 120 via said hub 107.

[0078] Preferably, said generator 120 is a permanent magnet synchronous generator.

[0079] Said generator 120 is connected to an active rectifier 130, which is connected to an inverter 140, which is connected to an electrical supply grid 250.

[0080] Said active rectifier 130 and said inverter 140 are preferably arranged as an back-to-back converter.

[0081] Said wind turbine 100 also comprises a wind turbine control unit (e.g., wind turbine controller) 150, which is designed to control said wind turbine 100, and in particular said active rectifier 130.

[0082] In order to control said wind turbine 100 said wind turbine control unit 150 may receive multiple signals, e.g., a measured phase current i.sub.g1, i.sub.g2, i.sub.g3 of said generator 120, a line voltage V between said active rectifier 130 and said inverter 140 and/or a power demand value P for controlling the power output of said wind turbine.

[0083] Said wind turbine control unit 150 is also designed to control said active rectifier 130 via a signal line in order to meet a specific operating point of said wind turbine or generator.

[0084] Moreover, said wind turbine control unit 150 is designed to control said active rectifier 130 using α/β-coordinates, e.g., for a current, in order to meet a specific operating point of said wind turbine and/or generator.

[0085] Said wind turbine control unit 150 may also comprise an optimization module 152 designed to run the herein-described optimization, e.g., as shown in FIG. 3.

[0086] FIG. 3 shows a flowchart 300 of a method for controlling an electric generator of a wind turbine, e.g., as shown in FIG. 1 and/or FIG. 2, comprising an optimization.

[0087] Said electric generator may be controlled via an active rectifier and said active rectifier is controlled via a power demand value P which is transformed into α/β-coordinates, which are used to control said active rectifier. Hence, said electric generator is controlled via a power demand value P using α/β-coordinates in order to drive an active rectifier controlling said electric generator.

[0088] In order to operate said generator properly also different operating parameters leading to different operating points are used, e.g., amplitude A and/or phase angle φ, in particular for controlling a generator current.

[0089] Any parameters/operating parameters needed for controlling said wind turbine may be provided by a wind turbine control unit 150, as shown in FIG. 2.

[0090] During operation of said wind turbine, an optimization 310 as described herein is carried out. Said optimization 310 is a mathematical optimization, in particular for operating parameters, e.g., amplitude A and phase angle φ, during which a suitable operating parameter A′, φ′ for controlling said wind turbine is determined, in particular in an iterative manner.

[0091] In the given example, the amplitude A and the phase angle φ are the operating parameters of the generator of the wind turbine.

[0092] In a first step 312 of the optimization, a multidimensional space is provided comprising a plurality of discrete operating points of said generator, e.g., as shown in FIG. 4A.

[0093] In a next step 314 of the optimization, an objective function for said multidimensional space is provided, e.g., a simplex or simplex-algorithm, in particular a Nelder-Mead method.

[0094] In a next step 316, the objective function 314 is applied to the multidimensional space 312 in an iterative manner for determining at least one parameter A, φ of said multidimensional space as a suitable operating parameter A′, (V.

[0095] Said suitable operating parameter A′, φ′ is then selected in a next step 320 and applied to a controller, e.g., for controlling 330 an active rectifier.

[0096] Thus, the wind turbine is controlled based on the selected suitable operating parameter A′, φ′ parameters as operating parameter A″, φ″.

[0097] In an embodiment, the suitable operating parameter A′, φ′ may also be used to update the multidimensional space, in particular the plurality of parameters Ai, φi.

[0098] FIG. 4A shows a multidimensional space 400, in particular as used in the optimization as shown in FIG. 3.

[0099] The multidimensional space 400 has multiple discrete points 412 and is in form of a two-dimensional polar coordinate system forming a grid of circular shape with approximately equal sized triangular cells.

[0100] Each discrete point 412 comprises a coordinate for the Amplitude A.sub.i and a coordinate for the phase angle φ.sub.i describing one operating point of said wind turbine, in particular the electric generator of said wind turbine.

[0101] By applying an objective function to said multidimensional space 400, a point 412 within said multidimensional space 400 may be sought having suitable parameters.

[0102] For determining whether parameters are suitable or not, a predetermined condition may be used, e.g., noise of the generator using a microphone with the hub of said wind turbine.

[0103] By using noise as a predetermined condition, the proposed method may be used to lower the noise of the generator and/or the wind turbine.

[0104] However, also other predetermined conditions may be used such as rotational speed of the generator, heat within said generator, mechanical disturbances, such as torque ripple, and further more.

[0105] FIG. 4B shows a multidimensional space 400 during optimization, in particular as shown in FIG. 3, using a predetermined condition, in particular noise.

[0106] The multidimensional space 400 is as shown in FIG. 4A.

[0107] In addition, the objective function OF is applied to said multidimensional space 400. This can be illustrated by the triangle 430.

[0108] Due to the objective function OF, a suitable parameter A′, φ′ is sought. This may be illustrated by the triangle 430 moving (arrow, 432).

[0109] In the given example, the noise of the generator is chosen as a predetermined condition. The noise of the generator is illustrated by lines d0, d1, d2, d3 each of the same noise.

[0110] The objective function will lead the triangle 430 to move into the target area TA referencing to the lowest noise of the generator. Hence, an Amplitude and a phase angle φ with the target area will be determined as a suitable parameter A′, φ′ leading to low noise of said generator.

[0111] FIG. 5 shows a flow chart 500 of the optimization 310, as shown in FIG. 3, in particular one embodiment of the step of: determining one parameter by applying said objective function to said multidimensional space, in particular in an iterative manner.

[0112] In this embodiment, a Nelder-Mead method is used to apply said objective function to said multidimensional space. Also, said Nelder-Mead method comprises a simplex having the shape of a triangle.

[0113] Hence, a numerical method is used to determine suitable, in particular optimal, operating parameter for controlling said wind turbine, in particular said electrical generator of said wind turbine.

[0114] In a first step 510, said optimization 310 is started.

[0115] In a next step 512, all points of the objective function are measured, in particular all three points of the triangle 430, as shown in FIG. 4B.

[0116] In a next step 514, a reflection of the worst point(s) of the measured points is performed. In this case, a reflection of the worst point of the triangle is performed according to the Nelder-Mead method.

[0117] Said reflection leads to a new triangle 520.

[0118] In a next step 522, it is checked whether a reference measurement is needed or not.

[0119] If so (y), said triangle is forced to the center point in a next step 524 and said center point is validated by measuring in a further step 526 or several furthers steps, e.g., by measuring five times. The amount of further steps depends on the allowed step size. After setting said center point, said optimization is started again 510.

[0120] If not so (n), and depending on the optimization mode, namely fast mode 600 or slow mode 700, new point(s) are measured.

[0121] In fast mode, it is checked if/whether said triangle does not move or not, in a first step 530. Said triangle is considered moving, if the last few triangle of the optimization have moved or not. If said last few triangle movements were not the same, only one new point is measured in further step 532. If said last few triangle movements were the same, the triangle is considered not moving and therefore, all points, in particular all three points, were measured in a next step 540.

[0122] In slow mode, all points, in particular all three points, were measured in a next step 540.

[0123] After measuring new point(s), it is checked weather or not a predetermined criteria is met, e.g., a threshold number of a counter, in a next step 550.

[0124] If said predetermined criteria is reached, a suitable operating parameter may be found and be selected as an operating parameter in a next step 560.

[0125] If said predetermined criteria is not reached, said reflection of the work points(s) of step 514 is repeated, followed by the same steps as described above and repeated until said predetermined criteria is met.

[0126] Therefore, said reflection and checking if said triangle is moving is performed in an iterative manner until said predetermined criteria is met.

[0127] Thus, said algorithm also comprises the steps of: measuring all needed points for the objective function. In this case: all three points of the triangle.

[0128] During optimization, different measurements are performed leading to new points and said new points may be stored in a look-up table (e.g., memory) 700. This is indicated by the (+) in the flow chart.

[0129] Said look-up table 700 may also comprise a counter 710, which is used if the predetermined criteria is a threshold number. This is indicated by the (#) in the flow chart.

[0130] Said look-up table 700 may also comprise a table 720 for an exponential moving average, which may be used to validate weather a parameter is suitable or not. This is indicated by the (*) in the flow chart.

[0131] FIG. 6 shows a look-up table 600, which is preferably used within the herein-described optimization.

[0132] Said loop-up table may be used to keep count of the measured points, as described above, in particular in FIG. 5.

[0133] Said look-up table 600 may comprises multiple tables 600′ each having a plurality of columns 602 and rows 604. The multiple tables 600′ are used for different temperatures Ti within said generator.

[0134] Said columns 602 are set for a first operating parameter and said rows 604 are set for a second operating parameter.

[0135] During operation the cells 610 of the look-up table 600 may be filled with measured values, e.g., noise, corresponding to a predetermined criteria, in particular by using the first and second operating parameter.

[0136] In one example, the first operating parameter is the amplitude and the second operating parameter is the phase angle, and the corresponding criteria is noise is measured via microphone in the nacelle.

[0137] Preferably, the cells 610 of said look-up table are update by using an exponential moving average (EMA).

[0138] More preferably, said operating parameters are binned.

REFERENCE CHARACTER LIST

[0139] 100 wind turbine [0140] 102 tower of said wind turbine [0141] 104 nacelle of said wind turbine [0142] 106 aerodynamic rotor of said wind turbine [0143] 107 hub of said wind turbine [0144] 108 rotor blade of said wind turbine [0145] 110 spinner of said wind turbine [0146] 120 generator, in particular a permanent magnet synchronous generator [0147] 130 active rectifier of said wind turbine [0148] 140 inverter of said wind turbine [0149] 150 wind turbine control unit of said wind turbine [0150] 152 optimization module [0151] 200 sensor, in particular accelerometer [0152] 250 electrical supply grid [0153] 300 scheme of the proposed method [0154] 310 run the optimization [0155] 320 selecting suitable operating parameters [0156] 330 controlling an active rectifier [0157] 400 multidimensional space [0158] 410 grid [0159] 412 discrete points of said grid [0160] 414 triangular cells [0161] 420 parameters [0162] 430 triangle [0163] A amplitude [0164] φ phase angle [0165] d0, d1, . . . lines of noise in dB [0166] TA target area [0167] Ti temperature [0168] i.sub.g1 generator current of a first phase [0169] i.sub.g2 generator current of a second phase [0170] i.sub.g3 generator current of a third phase [0171] y yes [0172] n no

[0173] The various embodiments described above can be combined to provide further embodiments. These and other changes can be made to the embodiments in light of the above-detailed description. In general, in the following claims, the terms used should not be construed to limit the claims to the specific embodiments disclosed in the specification and the claims, but should be construed to include all possible embodiments along with the full scope of equivalents to which such claims are entitled. Accordingly, the claims are not limited by the disclosure.