METHOD FOR CONTROLLING AN ELECTRIC GENERATOR OF A WIND TURBINE

Abstract

Provided is a method for controlling an electric generator of a wind turbine. The method includes varying an amplitude and/or phase angle of an harmonic current of said electric generator, while said electric generator is rotating, in particular at a known condition, measuring a signal indicative of generator vibration after and/or during varying said amplitude and/or phase angle, repeating said varying and said measuring until a predetermined requirement is met, evaluating an operating point for said electric generator by using said measured signals indicative of generator vibration in order to reduce a ripple torque of said generator, and controlling a current, in particular said harmonic current, of said electric generator in order to meet said operating point.

Claims

1. A method for controlling an electric generator of a wind turbine, comprising: varying an amplitude and/or a phase angle of a harmonic current of said electric generator, while said electric generator is rotating; measuring a signal indicative of a generator vibration after and/or during varying said amplitude and/or phase angle; repeating said varying and said measuring until a predetermined requirement is met; evaluating an operating point for said electric generator using said measured signals indicative of the generator vibration to reduce a ripple torque of said electric generator; and controlling a current of said electric generator to meet said operating point.

2. The method for controlling the electric generator according to claim 1, wherein said amplitude or phase angle is varied using a converter connected to said electric generator.

3. The method for controlling the electric generator according to claim 1, wherein said amplitude or phase angle are calculated using a counter ripple torque module.

4. The method for controlling the electric generator according to claim 1, wherein varying the amplitude and/or phase angle of the harmonic current including injecting a current.

5. The method for controlling the electric generator according to claim 1, wherein said signal indicative for the generator vibration is measured by a sensor placed on a hub or other structural part of said wind turbine.

6. The method for controlling the electric generator according to claim 1, wherein said predetermined requirement is a number of iteration cycles.

7. The method for controlling the electric generator according to claim 1, wherein said repeating of the varying and measuring is performed sequentially until said predetermined requirement is met.

8. The method for controlling the electric generator according to claim 1, wherein said repeating of the varying and measuring is performed by varying an offset to said phase angle or said amplitude.

9. The method for controlling the electric generator according to claim 1, wherein said operating point is evaluated to reduce vibrations of the 6.sup.th harmonic or a multiple of said 6.sup.th harmonic of said electric generator.

10. The method for controlling the electric generator according to claim 1, wherein said wind turbine further comprises a wind turbine controller and said method further comprises: implementing new parameters in said wind turbine controller based on said operating point.

11. The method for controlling the electric generator according to claim 10, wherein said new parameters include at least a torque set point.

12. The method for controlling the electric generator according to claim 1, wherein said operating point is generated using an adaptive or self-learning controller and/or by offset variation.

13. The method for controlling the electric generator according to claim 1, wherein said evaluating is performed based on a lookup table or by feedback loop control.

14. The method for controlling the electric generator according to claim 1, wherein: said evaluating includes a numerical method, an objective function, a simplex, a direct search method, a Nelder-Mead method and/or data binning; and/or said operating point is generated by a numerical method, an objective function, a simplex, a direct search method, a Nelder-Mead method and/or data binning.

15. The method for controlling the electric generator according to claim 1, wherein: said evaluating of the operating point includes a continuous process using a numerical method, an objective function, a simplex, a direct search method, a Nelder-Mead method and/or data binning; and/or said operating point is generated by a continuous process using a numerical method, an objective function, a simplex, a direct search method, a Nelder-Mead method and/or data binning.

16. The method for controlling the electric generator according to claim 1, wherein: said operating point is represented by at least one or more mechanical states of said electric generator; and/or said operating point includes at least one information of: a rotational position of a rotor of said electric generator; a temperature of said electric generator; a speed of said electric generator; a current of said electric generator; or a torque of said electric generator.

17. A method for operating a wind turbine, comprising the steps of: rotating an electric generator of said wind turbine at a known rotational speed and/or known torque; controlling the electric generator by at least: varying an amplitude and/or a phase angle of a harmonic current of said electric generator, while said electric generator is rotating; measuring a signal indicative of a generator vibration after and/or during varying said amplitude and/or phase angle; repeating said varying and said measuring until a predetermined requirement is met; evaluating an operating point for said electric generator using said measured signals indicative of the generator vibration to reduce a ripple torque of said electric generator; and controlling a current of said electric generator to meet said operating point; and feeding electrical energy into an electrical supply grid, while said electric generator is in said operating point.

18. The method for operating the wind turbine according to claim 17, wherein controlling the electric generator is started in response to a first criterion being met.

19. The method for operating the wind turbine according to claim 18, wherein said first criterion is checked by a controller.

20. The method for operating the wind turbine according to claim 17, wherein controlling the electric generator in response to a conditional criterion being met.

21. The method for operating the wind turbine according to claim 20, wherein said conditional criterion is checked by a wind turbine controller.

22. The method for operating the wind turbine according to claim 17, comprising: suppressing controlling the electric generator in response to a sensor detecting that the generator vibration is greater than a threshold or in response to the time being between 10 p.m. and 6 a.m.

23. A wind turbine, comprising: the electric generator, which is controlled by the method according to claim 1.

24. The wind turbine according to claim 23, comprising: a sensor placed on a hub of said wind turbine.

25. The wind turbine according to claim 24, wherein said sensor is configured to detect a vibration or a sound of said wind turbine, said electric generator and/or a wind turbine blade.

26. The wind turbine according to claim 24, wherein said sensor is an outdoor microphone or an accelerometer.

27. The wind turbine according to claim 23, further comprising: a wind turbine controller having a counter ripple torque module configured to calculate an amplitude or a phase angle of said electric generator.

28. The wind turbine according to claim 23, comprising: a converter coupled to said wind turbine controller and said electric generator and configured to control the current of said electric generator.

29. The wind turbine according to claim 23, comprising: an interface configured to: receive information used for evaluating said operating point; and/or send information indicating said evaluation to a third party interface.

30. The wind turbine according to claim 23, wherein said electric generator is a permanent magnet synchronous having an active rectifier configured to control the current of said electric generator.

31. The wind turbine according to claim 23, comprising: a controller configured to determine whether a criterion for performing said evaluating of the operating point is met.

32. The wind turbine according to claim 23, comprising: an adaptive controller configured to perform an offset variation method to obtain the operating point.

33. The method for controlling the electric generator according to claim 1, wherein controlling the current of the electric generator includes controlling the harmonic current.

34. The method for controlling the electric generator according to claim 2, wherein said amplitude or phase angle is varied using an active rectifier connected to said electric generator.

35. The method for controlling the electric generator according to claim 4, wherein varying the amplitude and/or phase angle of the harmonic current including injecting a 6.sup.th or 12.sup.th harmonic relative to a fundamental current frequency.

Description

BRIEF DESCRIPTION OF THE SEVERAL VIEWS OF THE DRAWINGS

[0093] With reference to the appended drawings, below follows a more detailed description of embodiments cited as examples.

[0094] FIG. 1 shows a wind turbine,

[0095] FIG. 2 shows another view of said wind turbine,

[0096] FIG. 3 shows a flowchart of a method,

[0097] FIG. 4A shows a lookup table which is used in the method.

[0098] FIG. 4B shows a lookup table which is used in the method, when a predetermined requirement is met.

[0099] FIG. 5 shows a continuous process using a Nelder-Mead method.

[0100] FIG. 6A shows a flowchart for a Nelder-Mead method.

[0101] FIG. 6B shows an advanced continuous process using binning.

[0102] FIG. 6C shows an example of an advanced bin table.

DETAILED DESCRIPTION

[0103] FIG. 1 shows a wind turbine 100.

[0104] Said wind turbine 100 comprises a tower 102 and a nacelle. Arranged on said nacelle 104 is an aerodynamic rotor 106 having a hub 107 with three rotor blades 108 and a spinner 110.

[0105] During operation, the rotor 106 is set in rotating motion by the wind and thereby drives a generator in said nacelle 104.

[0106] Said generator is preferably a permanent magnet synchronous generator with an active rectifier, which is designed to control a current of said generator.

[0107] 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.

[0108] Said sensor 200 said sensor is designed to detect a vibration or a sound of said wind turbine 100 and/or said electric generator and/or a wind turbine blade 108.

[0109] FIG. 2 shows another view of said wind turbine 100, in particular as shown in FIG. 1, in a preferred embodiment.

[0110] Said rotor blades 108 are mechanically connected to said generator 120 via said hub.

[0111] 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.

[0112] Preferably, said generator 120 is a permanent magnet synchronous generator.

[0113] 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.

[0114] Said wind turbine control unit 150 also comprises a counter ripple torque module 152 for calculating an amplitude or a phase angle of said electric generator 120. For this, said counter ripple torque module 152 is connected to said accelerometer 200, which is placed on an inner side of said hub 107 next to a pitch bearing.

[0115] In order to control said wind turbine 100 said wind turbine control unit 150 receives multiple signals, e.g., a measured phase current I1, I2, I3 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.

[0116] Moreover, said wind turbine control unit 150 is designed to control said active rectifier 130 via a current signal line in order to meet a specific operating point of said wind turbine or generator.

[0117] Said wind power control unit 150 may also comprise an interface, which is designed to receive an information, in particular a script, used for the evaluation of said operating point and/or to provide an information about said evaluation to a third party interface.

[0118] FIG. 3 shows a flowchart 300 of the proposed method, in particular a method for operating a wind turbine comprising the method for controlling an electric generator.

[0119] In a first step 310, the electric generator of said wind turbine is rotating at a known condition, in particular due to normal operations mode of said wind turbine. Therefore, said electric generator is running in a first operating point. This is shown by block ROT.

[0120] During the first step 310, it is checked whether the proposed method for controlling said electric generator needs to be started or not, e.g., by measuring the noise of said wind turbine. This is shown by block CHE.

[0121] If said noise is too much, said method for controlling an electric generator is started. For this, a predetermined noise limit may be used.

[0122] In a second step 320, the proposed method for controlling an electric generator is performed. In a first step of this method 321, an amplitude or phase angle of said electric generator is varied, while said electric generator is rotating, in particular at said known condition. This is shown by block VAR.

[0123] In a next step of this method 322, a signal indicative for a generator vibration is measured, in particular after varying said amplitude or phase angle. This is shown by block MEA.

[0124] The measured signals or corresponding values may be stored in a lookup table 330, by a step of storing 323.

[0125] These steps are repeated multiple times, until a predetermined requirement is met, by a step of repeating 324. For example, the amplitude is varied with each step until a predefined lookup table is fully filled with values.

[0126] In a next step of this method 325, in particular when said predetermined requirement is met, an operating point for said electric generator is evaluated by using said measured signals indicative for a generator vibration in order to reduce a ripple torque of said generator. For example, the amplitude providing the lowest noise is taken.

[0127] This value may also be used as a new setting for the generator and also be saved in the settings of said wind turbine, e.g., by the step of saving the evaluated value 326.

[0128] Afterwards, a current of said electric generator is controlled based on the new setting and/or the evaluated operating point in order to meet said operating point, e.g., by the step of controlling 327.

[0129] In the last step 330, electrical energy is fed into an electrical supply grid, while said generator is in said new operating point, namely a second operating point. In this operating point, the wind turbine produces less noise then in the first operating point. This is shown by block FEE.

[0130] Hence, a method is provided, wherein an offset variation is executed until enough data is collected. Afterwards, the data is evaluated for each bin and it is checked for the lowest offset, leading to number of points. To make this method more robust, a curve fitting is used, e.g., by using a least-square-method.

[0131] FIG. 4A shows a lookup table which is used in a preferred embodiment of the proposed method.

[0132] The lookup table 400 comprises two parts, namely a result matrix 410 and a corresponding counting matrix 420.

[0133] In both matrices 410, 420, the first row 412 is defined by fixed values, in particular the torque over the nominal torque of said electric generator in percent and the first column 414 is defined by fixed values, in particular the offset variation of the phase angle in percent.

[0134] In a first step of said method, said phase angle is varied and a specific torque is chosen.

[0135] In the shown example, 45% torque/nominal torque and −0.1% phase angle variation.

[0136] For this set of values, a signal or value is measured indicating the torque ripple of said electric generator.

[0137] The measured value is saved into said result matrix 410 and the counter for this of said counting matrix is set “+1.”

[0138] In a next step, e.g., after 3 rotations of said electric generator, said phase angle is varied again and a specific torque is chosen again.

[0139] Accordingly, the next measured value is then also saved in said result matrix 410 and the counter for this of said counting matrix is again set “+1.”

[0140] In the result matrix 410, each value is saved as a mean, for example via the equation:

[00001]$m_n=m_{n-1}+\frac{a_n-m_{n-1}}{n}.$

[0141] This process is repeated until a predetermined requirement is met, for example, five values collected for every option of said look up table. Such a lookup table is shown in FIG. 4B.

[0142] FIG. 4B shows a lookup table 400, which is used in a preferred embodiment of the proposed method, when a predetermined requirement is met.

[0143] Said counting matrix 420 met the predetermined requirement, namely five collected values for every option.

[0144] Therefore, the result matrix 410 comprises different values indicating the noise of said generator, wherein the lowest value represents the lowest noise of said generator.

[0145] According to the result matrix 410, the generator would produce the lowest noise at a specific phase angle 416, namely one which is −0.1% smaller than the current setting, at specific torque 418, namely 40 percent of the nominal torque of said generator.

[0146] In a preferred embodiment, in particular when the wind turbine is a torque controlled wind turbine, the results are saved in the settings so that the wind turbine control unit may choose the phase angle with the lowest value in accordance with the needed torque for proper operation.

[0147] For example, if a torque of 60% of the nominal torque is needed, the value 0.5 in the last column would be the lowest and therefore, a −0.1% smaller phase angle would be chosen.

[0148] During operation, the values may shift and therefore, the proposed method for controlling an electric generator may be repeated the next day or the next weak, or when said sensor is measuring a high noise, in particular too high noise.

[0149] FIG. 5 shows a continuous process 500 in a preferred embodiment, using a Nelder-Mead method.

[0150] Said continuous process 510 comprises three steps 512, 514, 516, which are looped.

[0151] In a first step 512, new parameters to be tested are selected and/or provided, in particular as described above.

[0152] In a next step 514, said new parameters are tested a number of times or until a predetermined requirement is met, as also described above.

[0153] Afterwards, the outcome of said test, in particular provided by step 514, is averaged and stored, e.g., in a look-up table.

[0154] Preferably, this may be done as long as the turbine operation is present.

[0155] In a more preferred embodiment, the status of said test is also updated. This is shown by step 516.

[0156] Once the continuous process 500 is finished, e.g., by meeting a predetermined requirement, the best settings may be stored or frozen in a look up table, e.g., a look-up table for a wind turbine or inverter controller, or any other database.

[0157] In a preferred embodiment, the best settings may also be converted into polar coordinates and stored as such.

[0158] Preferably, said continuous process 500 is used for evaluating an operating point for said electric generator, as described above or below.

[0159] FIG. 6A shows a flowchart 600 for a Nelder-Mead method, in particular as used in the continuous process shown in FIG. 5.

[0160] In a first step 610, said method is started, e.g., by building a simplex using points in a Cartesian grid. For this, the basics of a classical Nelder-Mead method may be used.

[0161] Starting said method may also comprise a step 612 for checking the age of said points. For this, time stamps may be used. For example, if the age of said points is too high and said points therefore fail the check (No), new points need be acquired in another step 614.

[0162] However, if the age of said points is low enough and said points therefore pass the check (Yes), said method proceeds as follows: [0163] In a next step 620, said points are sorted, e.g., the best point p.sub.l having the lowest vibration, the second best point p.sub.s and so on until the worst point p.sub.h having the highest vibration. [0164] In a next step 630, a reflection point p.sub.r is calculated based on the best point p.sub.l, the second best point p.sub.s and the worst point p.sub.h, e.g., by using the basics of a classical Nelder-Mead method.

[0165] Afterwards, several conditions are checked leading to different steps, in particular: [0166] if reflection point p.sub.r<best point p.sub.l is true, an expansion point p.sub.e is calculated in a next step 632, [0167] if reflection point p.sub.r>second best point p.sub.s is true, a contraction point p.sub.c is calculated in a next step 634.

[0168] Based on the contraction point p.sub.c several other conditions are checked leading to different steps, in particular: [0169] if contraction point p.sub.c<best point p.sub.h is true, it may be checked whether an allowed simplex minimum size is met in a next step 640. [0170] if contraction point p.sub.c≥best point p.sub.h is true, the points may be updated.

[0171] In a preferred embodiment, said method may also comprises a step of retesting 650, e.g., if the best point p.sub.h, the second best point p.sub.s or the worst point p.sub.l has changed.

[0172] Moreover, several security algorithms 602 may be implemented, e.g., for checking the maximum amplitudes.

[0173] Once the continuous process 600 is finished, e.g., by meeting a predetermined requirement, the best settings may be stored or frozen in a table, e.g., a look-up table for a wind turbine or inverter controller, or any other database.

[0174] In a preferred embodiment, the best settings may also be converted into polar coordinates and stored as such.

[0175] FIG. 6B shows an advanced continuous process 500′ using binning.

[0176] In this embodiment, said continuous process 500 uses binning leading to an advanced continuous process 500′.

[0177] In particular, the wind turbine operation is binned on a turbine variable or a combination of turbine variables, within each bin a simplex will operate, in particular as shown in FIG. 6A.

[0178] Therefore, the continuous process, as shown in FIG. 6A, is used for different bins. Said bin may be predetermined and/or set by the turbine manufacturer. In particular, the method of binning is used to optimize the used Nelder-Mead method even further.

[0179] In particular, whenever the operating point of the wind turbine shifts into a new bin, said bin is chosen and updated.

[0180] In a preferred embodiment, the simplex will be updated whenever the turbine operation enters a bin, in particular whenever the operating point enters a bin. This is shown by the step 518.

[0181] For example, the generator current is binned between 400 and 5500 ampere with 50 equidistant bins and outside these bins no or default counter steering is applied, which is shown in the bin table 710. Whenever the operating point shifts into another bin, the simplex will be updated for this bin as show in step 514, e.g., by testing several times.

[0182] FIG. 6C shows an example of an advanced bin table 700′, which may also be used in a method as shown in FIG. 6A or FIG. 6B.

[0183] The advanced bin table 700′ is multidimensional.

[0184] For example, the generator current is binned between 400 and 5500 ampere with 50 equidistant bins, as already shown by the bin table 700′ in FIG. 6B.

[0185] In addition, the rotor position is also binned, in particular between zero and 360 degrees with 30 equidistant bins, leading to a 2D bin.

[0186] In this example, the amplitude of the generator current and the mechanical angle of the rotor position are binned and used to find optimal counter steering parameters for the binned turbine operation.

[0187] For example, the same amplitude and the same phase angle are tested multiple times, e.g., 20 times, and averaged thereafter.

[0188] Due to the change of the mechanical angle of the generator during the test, the vibration may vary and therefore, the building of an average may counteract distortions within the test, leading to better results.

[0189] The method described herein is, in particular, used to measure harmonics of a generator and counter steer dominated harmonic thereof.

[0190] Furthermore, for each generator harmonic an own collection of binned simplexes may be used.

[0191] Moreover, multiple sensors may be used to obtain signals indicative for generator vibrations. Said signals may also be weighted, e.g., by a weighting function.

[0192] For longer measurement intervals, it might be beneficial to predict the upcoming operational point for choosing the right bin with the resulting optimization parameters.

REFERENCE CHARACTER LIST

[0193] 100 wind turbine [0194] 102 tower of said wind turbine [0195] 104 nacelle of said wind turbine [0196] 106 aerodynamic rotor of said wind turbine [0197] 107 hub of said wind turbine [0198] 108 rotor blade of said wind turbine [0199] 110 spinner of said wind turbine [0200] 120 generator, in particular a permanent magnet synchronous generator [0201] 130 active rectifier [0202] 140 inverter [0203] 150 wind turbine control unit [0204] 152 counter ripple torque module [0205] 200 sensor, in particular accelerometer [0206] 250 electrical supply grid [0207] 300 scheme of the proposed method [0208] 310 rotating generator at known condition [0209] 320 step of performing a method for controlling an electric generator [0210] 321 step of varying an amplitude or phase angle [0211] 322 step of measuring an indicative signal [0212] 323 step of storing the measured signal [0213] 324 step of repeating said varying and measuring [0214] 325 step of evaluating an operating point [0215] 326 step of saving [0216] 327 step of controlling [0217] 330 lookup table [0218] 331 settings [0219] 340 feeding electrical energy into an electrical supply grid [0220] 400 lookup table [0221] 410 result matrix [0222] 412 predefined row [0223] 414 predefined column [0224] 420 counting matrix [0225] 500 continuous process [0226] 500′ continuous process in a preferred embodiment [0227] 512 step of: new parameters [0228] 514 step of: testing new parameters [0229] 516 step of: updating parameters [0230] 518: step of: entering new bin and selecting new simplex [0231] 600 flowchart for a Nelder Mead method [0232] 602 step of: checking amplitudes [0233] 610 step of: starting [0234] 612 step of: checking simplex point age [0235] 614 step of: acquiring new points, [0236] 620 step of: sorting points [0237] 630 step of: calculating reflection [0238] 632 step of: calculating expansion point [0239] 634 step of: calculating contraction point [0240] 640 step of: checking simplex size [0241] 650 step of: retesting [0242] 700 bin table [0243] 700′ advanced bin table [0244] p.sub.h distinctive point with the highest vibration [0245] p.sub.l distinctive point with the lowest vibration [0246] p.sub.s distinctive point with the second lowest vibration [0247] pr reflection point [0248] p.sub.e expansion point [0249] p.sub.c contraction point