• 专利附录
  • 专利说明
  • 权利要求
  • 类似技术
  • 同族专利
  • 引用文献
  • 法律状态
  • 优先权
    • 专利摘要:
    The invention relates to a method of identifying a wind distribution pattern over a rotor plane and a wind turbine thereof. At least one operating parameter of the wind turbine and a rotational position of the rotor are measured over a time period. A first wind turbine blade passing signal is extracted from the measured operating parameter and a second wind turbine blade passing signal is generated from the rotational position. The first and second wind turbine blade passing signals are then analysed to determine the characteristics of the actual wind turbine blade passing signal in the rotor plane. These characteristics are afterwards compared to the characteristics of a plurality of known wind distribution patterns, and a unique relationship between the characteristics of the wind turbine blade passing signal and the wind distribution pattern is used to identify a distinctive wind distribution pattern.
    公开号:DK201670813A1
    申请号:DKP201670813
    申请日:2016-10-14
    公开日:2017-12-11
    发明作者:Lei Tong；Lin Chen；Peng Wang；Keld Stefan Pedersen
    申请人:Envision Energy (Jiangsu) Co Ltd；
    IPC主号:
    专利说明:

    Method of identifying a wind distribution pattern over the rotor plane and £ wind turbine thereof
    Field of the Invention
    The present invention relates to a method of identifying a wind distribution patterr over the rotor plane of a wind turbine and a wind turbine thereof. The method comprises measuring at least one operating parameter of the wind turbine, extracting c wind turbine blade passing signal from the measured operating parameters, and usim a unique relationship between characteristics of the wind turbine blade passing signal and a number of predetermined wind distributions to determine the wind distributior pattern over the rotor plane.
    Background of the Invention
    It is known that wind farms can be installed at locations having a flat and plain terrair where the wind is able to pass more or less freely over the terrain turbines. The wine turbines are thus subjected to a substantially undisturbed wind profile over the rotoi plane. It is also known that wind farms, due to various practical reasons, can be installed at locations having a complicated or very rough terrain, such as mountainous sites. Wind turbines installed at such sites tend to be subjected to unusual and harmful wind distributions over the rotor plane, such as extreme shears or reversed shears This results in unexpected aerodynamic imbalances over the rotor and, thus, accelerated accumulation of fatigue loads, which in turn leads to an increased risk of extreme loads and unexpected accidents, and also reduced operating life time for the wind turbine. One way to solve these problems is to provide a wind turbine control systeir having a load protection algorithm, thereby allowing the wind turbine to reduce loadf in critical situations, e.g. in the event of extreme loads. However, this load protective mode affects the total power production of the wind turbine. WO 2013/182204 A1 discloses a control method for minimising asymmetrical loac moments acting on the rotor where sensors directly or indirectly measure the mechanical loads. The measured signals are processed to determine a tilt moment and a yaw moment which are then analysed to determine a 3P-component thereof. This 3P component and a reference value are used to calculate an error value which, in turn, if used to determine a 2P-cyclic pitch value for each wind turbine blade. The individua pitch angles of the respective wind turbine blades are then adjusted using this 2P- cyclic pitch value in order to compensate for the asymmetrical load moments. Thii control system fails to disclose or suggest how to identify and recognise wind distribution patterns over the rotor plane. WO 2010/016764 A1 discloses alternative solution for minimising the asymmetrical load moments, wherein the bending moment of the rotor shaft is determined based or the sensor signals. Suitable corrective pitch values for each wind turbine blade are then calculated based on the bending moment components using an inverse trigonometric relationship between the deflection moment of the wind turbine blade and the bending moment of the rotor axis. The individual pitch angles are then adjusted according to these corrective pitch values. This control system also fails to disclose oi suggest how to identify and recognise wind distribution patterns over the rotor plane. EP 2025929 B1 discloses a predictive control method where the wind speed, wine shear, and wind turbulence in front of the rotor plane are measured by means of a LI-DAR system. The measured wind data is converted into predicted aerodynamic loac signals and then compared with measured load signals of suitable sensors located ir the wind turbine. The difference is then used to calculate a corrective pitch value which, in turn, is applied to each respective wind turbine blade in order to reduce the asymmetrical loads in advance and thus maintain a uniform load on the rotor. Thii control scheme requires a complex measurement of the wind profile at a distance frorr the rotor, however, this wind profile may change before reaching the rotor and thuf changing the aerodynamic loads. Secondly, such a hub mounted LIDAR system addf to total costs of the wind turbine. Thirdly, it is not hinted in EP 2025929 B1 that the measured wind profile or load signals can be used to identify and recognise wind distribution patterns over the rotor plane. WO 2015/192856 A1 discloses a control method where integrated strain gauges oi fibre optical sensors measure the mechanical loads on the rotor. A tilt moment and £ yaw moment of the main bearing are calculated based on the load signal from the sensors. The tilt and yaw moments and the measured wind speed are used to estimate £ horizontal and a vertical wind shear over the rotor plane. The horizontal and vertical wind shears are then used to estimate a wind velocity value which, in turn, is used tc calculate an optimal tip speed ratio and an optimal pitch angle. This optimal tip speec ratio and optimal pitch angle are then applied to the respective wind turbine blades. Ir this control scheme, it is stated that the power production is optimised and the costs o] energy of the wind turbine is reduced by adjusting the rotational speed and pitch angle of the wind turbine blades. It is further stated that this solution is only suitable for higl wind shear conditions. It is not hinted that the estimated wind shears are used to identify and recognise different wind distribution patterns over the rotor plane, nor tha they are stored for later root cause analysis.
    Direct measurement of the wind distribution over the rotor plane is very difficult tc achieve and is thus often not available to the wind turbine operator. Therefore, there exists a need for a method of identifying the wind distribution over the rotor plane under ambient conditions which allows for root cause analysis as well as protective control of the wind turbine.
    Object of the Invention
    An object of this invention is to provide a method of recognising and identifying the wind distribution pattern over the rotor plane.
    Another object of this invention is to provide a method of determining the wind distribution over the rotor plane.
    Yet another object of this invention is to provide a method that allows for later roo cause analysis based on measured wind distribution conditions. A further object of this invention is to provide a wind turbine with a control systerr that allows for the reduction of extreme loads in the wind turbine.
    Description of the Invention
    An object of the invention is achieved by a method of identifying a wind distributior pattern over a rotor plane of a wind turbine, the wind turbine comprising a rotor whict is rotary arranged relative to a nacelle, the nacelle is arranged on top of a wind turbine tower, the rotor comprises at least two wind turbine blades mounted to a rotatable hub wherein the at least two wind turbine blades define the rotor plane, the wind turbine further comprising an angular sensor configured to measure a rotational position, a least a second sensor configured to measure at least one operating parameter, and c control system electrically connected to the angular sensor and the at least second sensor, wherein the method comprises the steps of: measuring at least one operating parameter, measuring a rotational position of the rotor, determining a wind distribution over the rotor plane, e.g. in at least a horizontal direction or a vertical direction, characterised in that, the method further comprises the steps of: comparing said wind distribution to a plurality of predetermined wind distribution patterns, and identifying a match between the wind distribution and one ol said plurality of predetermined wind distribution patterns.
    This provides a simple and easy method of identifying a wind distribution pattern ovei the rotor, such as the prevailing wind distribution pattern, the most frequently occurring wind distribution pattern, or other relevant wind distribution patterns. This method may also be used to identify changes in the wind distribution patterns. The presem method eliminates the need for a hub mounted LIDAR system, since the method maj use the already existing sensor signals found in the wind turbine. The present methoc also allows for an improved protective action or improved root cause analysis. The present method can suitable be used for any type of variable speed wind turbines. Furthermore, the present method can suitable be used in a remote or local control systeir wherein the present method can be implemented as a control algorithm in the remote control system or in the local control system. The control system may be a remote oi local control system configured to control the operation of the variable speed wine turbine. The method can suitably be used at all wind shear conditions, i.e. at low wine shears, at medium wind shears, as well as high wind shears.
    The rotational position is measured as the angular position of the wind turbine bladef in the rotor plane relative to a reference angle, e.g. a vertical position where the tip end of the wind turbine blade faces away from the wind turbine tower. The rotationa speed of the rotor also defines a rotational frequency, i.e. IP frequency, of the rotor.
    The rotational frequency may be used to calculate a passing frequency of the wine turbine blades, e.g. the 2P or 3P frequency, by multiplying the rotational IVequenc with the number, e.g. two, three, or more, of wind turbine blades. This wind turbine blade passing frequency may be used to determine the actual phase of the wind turbine blade passing signal in the rotor plane as described later.
    According to one embodiment, said step of determining a wind distribution over the rotor plane comprises: determining at least one wind turbine blade passing signal based on said a least one operating parameter or said rotational position, calculating the characteristics, e.g. an amplitude or a phase, of said one wine turbine blade passing signal.
    The operating parameter, the rotational speed, or another suitable signal may be usee to determine the characteristics of the wind turbine blade passing signal in the rotoi plane. The characteristics may be defined by the amplitude and at least the phase ol the actual wind turbine blade passing signal. In example, the operating parametei and/or the rotational position may be suitably processed, e.g. filtered and/or amplified to determine the actual wind turbine blade passing signal. This signal may then be suitably analysed, e.g. in the time and/or frequency domain, to determine the characteristics, e.g. the phase and/or the amplitude. This eliminates the need for additiona sensors or sensor units, since the wind turbine blade passing signal can be determinec by using the signal of the sensors already situated in the wind turbine. This also reduces the total costs of implementing the present method in existing wind turbines or ir new wind turbines.
    In example, the operating parameter may be analysed to determine at least the amplitude of the actual wind turbine blade passing signal. In example, the rotational position may be analysed to determine the phase of the actual wind turbine blade passing signal.
    According to a special embodiment, each of the predetermined wind distribution patterns is defined by a unique relationship between said each wind distribution patterr and predetermined characteristics of the wind turbine blade passing signal, whereir said step of comparing said wind distribution to a plurality of predetermined wine distribution patterns comprises comparing the calculated characteristic to said predetermined characteristics.
    Conventional control methods use the measured load signals, e.g. the 3P loads, to directly determine a corrective pitch angle or tip speed ratio used to perform a cyclic pitching of the wind turbine blades. No further analysis of the measured load signals if performed in order to identify different wind distribution patterns over the rotor plane.
    In example, a unique relationship may be established between a distinctive wind distribution pattern and distinctive characteristics of the wind turbine blade passing signal. This unique relationship may be established using simulations, wind shear models, previous wind distribution measurements, or other suitable means. Each wind distribution pattern is defined by a set of horizontal and vertical wind shear values representing the x-axis and z-axis in the rotor plane. The y-axis is defined by the rotatior axis of the rotor shaft. A first set of distinctive wind distribution patterns may be established for a clockwise rotation of the rotor. A second set of distinctive wind distribution patterns may be established for an anti-clockwise rotation of the rotor. Thif unique relationship allows the control system to determine a current wind distributior pattern over the rotor plane.
    In example, the unique relationship may further be used to recognise and identify different wind distribution patterns over the rotor plane. The current wind distributior pattern may be compared to at least one set of known wind distribution patterns, e.g the first or second set mentioned above in order to establish a match or at least a substantial match. This match or substantial match between the current wind distributior pattern and a distinctive wind distribution pattern may be achieved by performing i weighted evaluation of the current wind shear values relative to the stored wind sheai values of the known wind distribution patterns. The weighted evaluation may optionally generate a relative value indicative of the likelihood of a match. If this relativt value exceeds a predetermined threshold, then a match is established. This identifiec wind distribution pattern can advantageously be used to determine a suitable protective action as described above or improve the root cause analysis if a failure or damage is detected, e.g. due to extreme or unexpected loads acting on the rotor.
    Alternatively, the currently calculated characteristics of the wind turbine blade passing signal may be compared to the individual characteristics of the known wind distribu tion patterns in order to establish a match or at least a substantial match. This match oi substantial match may be established in the same manner as described above or b using a different technique.
    Optionally, the current wind distribution pattern may be determined by applying £ wind shear transfer function to the wind turbine blade passing signal, e.g. to the calculated amplitude and/or phase.
    According to one embodiment, at least the at least one operating parameter or the rotational position of the rotor is measured over a predetermined time period.
    The operating parameter, the rotational speed, and said another suitable signal may be measured over a predetermined time period. The time period may suitably be selectee dependent of the particular configuration of the wind turbine and/or the particular application (load reduction control or root cause analysis) of the present method. In example, but not limited to, the respective signals may be measured over a relative shor time period, e.g. up to 10 minutes, e.g. between 1 minute and 5 minutes. In example but not limited to, the respective signals may be measured over a relatively long time period, e.g. up to 5 hours, e.g. between 5 minutes and 2 hours. The measured signal oi signals may then be suitably processed, e.g. using statistical analysis, to generate the wind turbine blade passing signal. This provides a more accurate measurement of the wind distribution and reduces the influence of sudden wind gusts.
    According to one embodiment, the method further comprises the step of updating the predetermined wind distribution patterns.
    The stored wind distribution patterns, e.g. the distinctive characteristics thereof, maj be updated, including adding new wind distribution patterns, deleting old wind distribution patterns, or updating existing wind distribution patterns. This may be achiever by applying a machine learning algorithm to the current wind distribution pattern oi the characteristics thereof, or by automatically or manually updating the stored wine distribution patterns using an internal or external database. Any suitable machine learning algorithm may be used to update the known wind distribution patterns. The allows for a more accurate recognition of distinctive wind distribution patterns. The also allows the wind distribution patterns to be adapted to the specific location of the wind turbine. This, in turn, may provide a more accurate load reduction control or roo cause analysis.
    The wind distribution patterns may be updated after each run of the present process, 01 when it is deemed necessary. Alternatively, the wind distribution patterns may be updated upon request, e.g. from a remote operator or external service provider.
    According to one embodiment, said step of identifying a match between the wind distribution and one of said plurality of predetermined wind distribution patterns comprises applying a pattern recognition algorithm to the calculated characteristics of the wind turbine blade passing signal.
    The identification of wind distribution patterns may comprise applying a patterr recognition algorithm to the wind turbine blade passing signal, e.g. the calculated amplitude and/or phase. Any suitable pattern recognition algorithm may be used to identify a distinctive wind distribution pattern. The pattern recognition algorithm may use one or more parameters or classes to determine a match or a substantial match. In example, the pattern recognition algorithm may determine a distribution or averagec value of the calculated amplitude and/or phase over the measured time period. This distribution or averaged value may then be used to identify a distinctive wind distribution pattern.
    These parameters or classes may further be updated, as mentioned above, to improve the values of the current wind distribution pattern and/or to improve the recognition oi a distinctive wind distribution pattern.
    According to one embodiment, said step of calculating the characteristics of said one wind turbine blade passing signal comprises calculating an amplitude and a phase, the amplitude and the phase being indicative of the wind turbine blade passing signal oi the rotor plane.
    As mentioned earlier, the operating parameter, the rotational speed, or said anothci suitable signals may be used to determine the amplitude and the phase of the actual wind turbine blade passing signal in the rotor plane. In example, a filter algorithm maj be applied to the operating sensor signal, i.e. the measured operating parameter, foi extracting a first wind turbine blade passing signal. The filter algorithm may in example, but not limited to, be a notch filter or a Kalman filter. This extracted signal maj then be transferred into the frequency domain, e.g. by applying a frequency transfei function. The frequency transfer function may in example, but not limited to, be a fasi Fourier transform (FFT) algorithm. This frequency transformed signal may be analysed to determine the amplitude, e.g. the relative amplitude, of the actual wind turbine blade passing signal.
    In example, a second wind turbine blade passing signal may be generated by multiplying the angular sensor signal, i.e. the measured rotational position, with the number ol wind turbine blades. This generated signal may then be analysed in the time domain e.g. by applying integral function, to determine a phase error relative to the first wine turbine blade passing signal. The integral function may in example, but not limited to be a correlation analysis algorithm or a Hilbert transformation algorithm. The correlation analysis algorithm may be an auto-correlation algorithm or a cross-correlatior algorithm. This phase error may then be used to determine the phase of the actua wind turbine blade passing signal.
    According to one embodiment, said at least one operating parameter is selected from £ generator torque signal, a generator speed signal, a rotor torque signal, a vibratior signal, or a blade bending moment signal.
    The operating parameter may be measured using various sensors capable of measuring an operating parameter of the wind turbine. The operating parameter may in example be a generator torque signal, a generator or rotor speed signal, a rotor torque signal, £ vibration signal, a blade bending moment signal, or another suitable operating signal The vibration signal may be measured in a fixed frame of reference, e.g. defined b the wind turbine tower, and along the x-axis and/or y-axis. The blade bending momen signal may be measured directly, or be measured as a load signal which, in turn, h used to calculate the blade bending moment. This allows the already available sensoi signals to be used as input for this present method.
    According to one embodiment, said method further comprises at least the step of: operating the wind turbine in a load protective mode, wherein the configuration of the wind turbine in said load protective mode is selected according tc the identified wind distribution pattern, or storing the wind distribution in a database, e.g. transmitting said stored wine distribution to a remote control system or monitoring unit.
    The present method can suitably be used for load reduction control, wherein the present method may be implemented in the control system of the wind turbine as a loac protective algorithm. The method may be configured to perform such load reductior control in real time, at regular time intervals, or upon detecting certain events, e.g detecting extreme loads. The wind turbine may be operated in a normal operatior mode, i.e. producing a maximum or nominal power output, or in at least one load protective mode, i.e. reducing loads on the wind turbine. In the load protective mode, c protective action may be applied in the event that certain loads and/or wind condition! are detected. This allows the aerodynamic or fatigue loads acting on the wind turbine to be reduced.
    The protective action may be applied to the wind turbine by adjusting one or more oj the control signals used to operate the wind turbine so that these loads are reduced The control signals may in example, but not limited to, be a generator torque contro signal, a power output control signal, a rotor speed control signal, a pitch angle control signal, or another suitable control signal. The protective action may be selected according to the identified wind distribution pattern. In example, a first protective actior may be applied when identifying a first wind distribution pattern, a second protective action may be applied when identifying a second wind distribution pattern, and so on This allows the protective action to be adapted in accordance with the current wine distribution pattern.
    Alternatively, the wind distribution patterns may be linked to individual protective actions where each individual protective action is defined by a set of control signals Each set of control signals may be optimised according to that particular wind distribution pattern and/or the particular configuration of the wind turbine in order to ar optimal load reduction control. These sets of control signals may be stored in the control system and used to operate the wind turbine in the load protective mode.
    The present method may also be used to perform a root cause analysis (RCA) wher detecting a failure or damage in or on a component of the wind turbine. The operating parameter, the rotational position and/or speed, the wind distribution patterns, anc other data may be stored in a database. These stored data may then be analysed at £ later time, e.g. using models, estimations, correlations, or other techniques, in order tc identify the reason of the failure or damage. The identification of distinctive wind distribution patterns provides an improved basis for performing the RCA and thus allowf for a more accurate identification of the reason for the failure or damage. The storec data may alternatively be transmitted to a remote wind turbine control system or £ remote monitoring unit. This allows the wind turbine operator to monitor the performance of the wind turbine, or compare the stored data with data collected from othci wind turbines located in that area.
    An object of the invention is also achieved by a wind turbine comprising: a rotor rotary arranged relative to a nacelle, the rotor comprises at least twc wind turbine blades mounted to a rotatable hub, wherein the at least two wine turbine blades define a rotor plane, the nacelle being arranged on top of a wind turbine tower, an angular sensor configured to measure a rotational position of the rotor, at least a second sensor configured to measure at least one operating parametei of the wind turbine, a control system electrically connected to the angular sensor and the at leas second sensor, characterised in that, the control system is configured to determine a wind distribution over the rotor plane, e.g. in at least a horizontal direction or a vertical direction, wherein the control system is further configured tc compare said wind distribution to a plurality of predetermined wind distribution patterns and to identify a match between the wind distribution and one oj said plurality of predetermined wind distribution patterns.
    This provides a wind turbine capable of identifying the wind distribution pattern acting on the rotor plane and optionally any changes in the wind distribution pattern. Thh is particularly relevant for wind turbines installed in complicated or very rough terrains, such as mountainous sites. The wind turbine may be any type of a variable speed wind turbine. This allows for an improved load protective control as well as ar improved RCA in the event of a failure or a detected damage. The present inventior may be integrated into new wind turbines or retrofitted into existing wind turbines.
    The rotational position is measured by at least one angular sensor, e.g. a rotary encoder. The angular sensor may include a calibration method or receive a calibration signal for compensating any drifts in the measured angular signal. The operating parametei is measured by at least one operating sensor. The operating sensor may in example be a torque sensor, a speed sensor, a power sensor, a pitch angular sensor, a vibratior sensor, a load sensor, or another suitable sensor. The vibration sensor, e.g. an accelerometer, may be configured to measure the vibrations in at least one direction, e.g along the x-axis and/or the y-axis. The load sensor, e.g. strain gauges, may be configured to be arranged relative to a component of the wind turbine, e.g. the wind turbine blade or the rotor shaft. The control system may comprise an optional database and £ microprocessor, a programmable logic controller (PLC), or another suitable controller The control system is electrically connected to the respective sensors and other electrical components of the wind turbine.
    According to one embodiment, said control system is configured to determine at leas one wind turbine blade passing signal from said at least one operating parameter οι said rotational position, and to calculate the characteristics, e.g. at least an amplitude or a phase, of said one wind turbine blade passing signal.
    The control system may be configured to extract a first wind turbine blade passing signal from the measured operating parameter, and to generate a second wind turbine passing signal from the measured rotational position. The control system may furthei be configured to determine the characteristics of the actual wind turbine blade passing signal in the rotor plane based on these two wind turbine blade passing signals as described earlier. The control system may additionally be configured to compare the current characteristics to the characteristics of a plurality of known wind distributior patterns in order to identify a match. This enables the control system to recognise anc identify a distinctive wind distribution pattern over the rotor plane. This wind distribu tion pattern may be used in a load protective mode or in a RCA. The use of wind turbine blade passing signals to identify wind distribution patterns allows for a simple and easy implementation of this method into a control system.
    According to a special embodiment, each of the predetermined wind distribution patterns is defined by a unique relationship between said each wind distribution patterr and predetermined characteristics of the wind turbine blade passing signal, whereir the control system is configured to compare the calculated characteristic to said predetermined characteristics for identifying the match. A unique relationship between the wind distribution pattern and the characteristics ol the wind turbine blade passing signal may be used to identify a current wind distribution pattern or any distinctive wind distribution patterns as described earlier. A plurality of known wind distribution patterns may be stored in a database connected to a controller. This database may be updated by a machine learning algorithm in the controller or via an external or internal database.
    The control system may be configured to store the characteristics of the actual wine turbine blade passing signal and/or the current wind distribution pattern in the database or in another database. The controller may be configured to control the communication with a remote control system or monitoring unit, and to transmit the storec data to this remote control system or monitoring unit.
    According to one embodiment, said control system is configured to operate the wine turbine in a normal operation mode and in at least one load protective mode, whereir the control system in the at least one load protective mode is configured to apply c protective action to the wind turbine based on the identified wind distribution pattern.
    The control system may be configured to operate a wind turbine in a load protective mode when detecting certain events, such as extreme loads or extreme wind shears. Ir this load protective mode, the controller may initiate a protective action by adjusting one or more of the control signals used to control the operation of the wind turbine ir a normal operation mode. The protective action may in example, but not limited to, be shutting the wind turbine down, derating the performance of the wind turbine, per forming an individual pitch control of the wind turbine blades, or another suitable protective action. When the control system detects that the event is no longer present then the control system may operate the wind turbine in the normal operation mode.
    The control system may alternatively be configured to select the required protective action based on the identified wind distribution pattern. This allows the control signah used to operate the wind turbine to be optimised according to each distinctive wine distribution pattern. The optimised values of these control signals may further be stored in the database where the control system may transmit these optimised control signals to the respective operating unit in the wind turbine. This allows for an improved load reduction control and, optionally, a reduced loss of power productior when operating in the load protective mode.
    The invention is not limited to the embodiments described herein, and thus the described embodiments can be combined in any manner without deviating from the objects of the invention.
    Description of the Drawing
    The invention is described by example only and with reference to the drawings wherein:
    Fig. 1 shows an exemplary embodiment of a wind turbine,
    Fig. 2 shows an exemplary graph of the measured operating parameter,
    Fig. 3 shows a first wind turbine blade passing signal extracted from the measurer operating parameter,
    Fig. 4 shows an exemplary graph of the measured rotational position,
    Fig. 5 shows an exemplary graph of the phase error between the actual wind turbine blade passing signal, the rotational position, and a second wind turbine blade passing signal, and
    Fig. 6a-c show the unique relationships between the characteristics of the wind turbine blade passing signal and three distinctive wind distribution patterns.
    In the following text, the figures will be described one by one, and the different parti and positions seen in the figures will be numbered with the same numbers in the dif- ferent figures. Not all parts and positions indicated in a specific figure will necessarih be discussed together with that figure.
    Position number list 1. Wind turbine 2. Tower 3. Nacelle 4. Rotor 5. Hub 6. Wind turbine blades 7. Control system 8. Sensor units 9. Operating parameter 10. First wind turbine blade passing signal 11. Rotational position 12. Second wind turbine blade passing signal 13. Actual wind turbine blade passing signal 14. First wind distribution pattern 15. Second wind distribution pattern 16. Third wind distribution pattern
    Detailed Description of the Invention
    Fig. 1 shows an exemplary embodiment of a wind turbine 1 in the form of a variable speed wind turbine. The wind turbine 1 comprises a wind turbine tower 2 provided or a foundation. A nacelle 3 is arranged on top of the wind turbine tower 2 and configured to yaw relative to the wind turbine tower 2 via a yaw system (not shown). A rotoi 4 comprising a hub 5 and at least two wind turbine blades 6 is rotatably arranged relative to the nacelle 3, wherein the wind turbine blades 6 are mounted to the hub 5. The wind turbine 1 is here shown with three wind turbine blades 6. The hub 5 is connecter to a drive train arranged in the nacelle 3 via a drive shaft, wherein the drive train comprising at least a generator for producing an electrical power output.
    The wind turbine 1 further comprises a control system 7 in the form of a local control system connected to a plurality of sensor units 8. The control system 7 is configured tc operate the wind turbine 1 in a normal operation mode and in at least one load protective mode. The sensor units 8 include an angular sensor configured to measure a rotational position of the rotor 4 and at least an operating sensor configured to measure ar operating parameter of the wind turbine 1.
    Fig. 2 shows an exemplary graph of the measured operating parameter 9 of a clockwise rotating rotor 4. The operating parameter is here measured as a generator speec signal using a suitable generator speed sensor. The x-axis indicates the measured time in milliseconds [ms] while the y-axis indicates the measured number of revolutions o] the generator in rounds per minute [rpm].
    Fig. 3 shows a first wind turbine blade passing signal 10, 10’ extracted from the measured operating parameter 9. The wind turbine blade passing signal is here measured as a 3P signal of the wind turbine 1 and is processed in the time and frequency domains respectively. The x-axis indicates the measured time in milliseconds [ms] anc in hertz [Hz] respectively while the y-axis indicates the relative amplitude of the signal. The amplitude is here shown in [rpm], but can also be measured in radians pei second [rad/s].
    The first wind turbine blade passing signal is here extracted by having applied a Kalman filter algorithm to the operating sensor signal, i.e. the measured operating parameter 9. This extracted signal 10 is further transformed into the frequency domain b applying a FFT-algorithm. This frequency transformed signal 10’ is then analysed tc determine the relative amplitude of the actual wind turbine blade passing signal.
    Fig. 4 shows an exemplary graph of the measured rotational position 11. The rotational position is here measured as an angular signal using a suitable angular sensor. The x-axis indicates the measured time in milliseconds [ms] while the y-axis indicates the angular position of the rotor 4 in degrees. The measured angular positions are here shown in multiples of 360 degrees, i.e. the angular position is rest to zero each time the angular position passes 360 degrees.
    Fig. 5 shows an exemplary graph of the phase error between the actual wind turbine blade passing signal, the rotational position 11 and a second wind turbine blade passing signal 12. The x-axis indicates the measured time in milliseconds [ms] while the y-axis indicates the relative angular position of the rotor 4 in degrees.
    The rotation speed 11 signal is here processed in the time domain by multiplying i with the number of wind turbine blades 6 to generate the second wind turbine blade passing signal 12. This second wind turbine blade passing signal 12 is then analyser using a Hilbert transformation algorithm to determine a phase error in relation to the first wind turbine blade passing signal 10. This phase error is then used to determine the phase of the actual wind turbine blade passing signal 13.
    Fig. 6a-c show the unique relationships between the characteristics of the actual wine turbine blade passing signal 13 and three distinctive wind distribution patterns 14, 15 16. In the upper diagram, the x-axis indicates the width of the rotor plane defined b the rotor 4 in metres [m] while the y-axis indicates the height of this rotor plane ir metres [m]. The reference point {0m, 0m} indicates the rotational axis or y-axis of the wind turbine 1. The wind speed is here measured in metres per second [m/s]. In the lower diagram, the radius indicates the relative amplitude of the actual wind turbine blade passing signal 13 while the radial extending lines indicate the phase of the actua wind turbine blade passing signal 13.
    Fig. 6a shows the unique relationship between a first distinctive wind distribution pattern 14 and the corresponding characteristics, e.g. the amplitude and phase, of the actual wind turbine blade passing signal 13’. The first distinctive wind distribution pattern 14 is here shown as vertical wind shears extending in an upwards direction, i.e parallel to the longitudinal direction of the wind turbine tower 2 and with the highesi wind speeds at the uppermost part of the rotor plane.
    Fig. 6b shows the unique relationship between a second distinctive wind distributior pattern 15 and the corresponding characteristics, e.g. the amplitude and phase, of the actual wind turbine blade passing signal 13”. The second distinctive wind distributior pattern 15 is here shown as vertical wind shears in opposite direction, i.e. parallel tc the longitudinal direction of the wind turbine tower 2 and with the highest wine speeds at the lowermost part of the rotor plane.
    Fig. 6c shows the unique relationship between a third distinctive wind distributior pattern 16 and the corresponding characteristics, e.g. the amplitude and phase, of the actual wind turbine blade passing signal 13’”. The third distinctive wind distributior pattern 16 is here shown as a combination of horizontal wind shears, i.e. perpendiculai to the longitudinal direction of the wind turbine tower 2, and vertical wind shears, e.g with the highest wind speeds at the lower-left part of the rotor plane.
    When the characteristics of the actual wind turbine blade passing signal 13 are determined, they are then compared to a plurality of stored distinctive wind distributior patterns stored in a database in order to identify a match. Each distinctive wind distribution pattern is defined by distinctive characteristics of the actual wind turbine blade passing signal 13 as shown in fig. 6a, fig. 6b, and fig. 6c. This enables the control system 7 to identify and recognise the distinctive wind distribution patterns acting on the rotor plane by simply analysing the wind turbine blade passing signal generated in the wind turbine 1.
    权利要求:
    Claims (13)
    [1] 1. A method of identifying a wind distribution pattern over a rotor plane of a wine turbine, the wind turbine comprising a rotor rotary arranged relative to a nacelle, the nacelle is arranged on top of a wind turbine tower, the rotor comprises at least twe wind turbine blades mounted to a rotatable hub, wherein the at least two wind turbine blades define the rotor plane, the wind turbine further comprising an angular scnsoi configured to measure a rotational position, at least a second sensor configured tc measure at least one operating parameter, and a control system electrically connectec to the angular sensor and the at least second sensor, wherein the method comprises the steps of: measuring at least one operating parameter, measuring a rotational position of the rotor, determining a wind distribution over the rotor plane, e.g. in at least a horizontal direction or a vertical direction, characterised in that, the method further comprises the steps of: comparing said wind distribution to a plurality of predetermined wind distribution patterns, and identifying a match between the wind distribution and one oj said plurality of predetermined wind distribution patterns.
    [2] 2. A method according to claim 1, characterised in that, said step of determining £ wind distribution over the rotor plane comprises: determining at least one wind turbine blade passing signal based on said a least one operating parameter or said rotational position, calculating the characteristics, e.g. an amplitude or a phase, of said one wine turbine blade passing signal.
    [3] 3. A method according to claim 2, characterised in that, each of the predeterminec wind distribution patterns is defined by a unique relationship between said each wine distribution pattern and predetermined characteristics of the wind turbine blade passing signal, wherein said step of comparing said wind distribution to a plurality of predetermined wind distribution patterns comprises comparing the calculated characteristic to said predetermined characteristics.
    [4] 4. A method according to any one of claims 1 to 3, characterised in that, the at leasi one operating parameter or the rotational position of the rotor is measured over a predetermined time period.
    [5] 5. A method according to any one of claims 1 to 4, characterised in that, the methoc further comprises the step of updating the predetermined wind distribution patterns.
    [6] 6. A method according to any one of claims 2 to 5, characterised in that, said step oj identifying a match between the wind distribution and one of said plurality of predetermined wind distribution patterns comprises applying a pattern recognition algorithm to the calculated characteristics of the wind turbine blade passing signal.
    [7] 7. A method according to any one of claims 2 to 6, characterised in that, said step ol calculating the characteristics of said one wind turbine blade passing signal comprises calculating an amplitude and a phase, the amplitude and the phase being indicative oj the wind turbine blade passing signal of the rotor plane.
    [8] 8. A method according to any one of claims 1 to 7, characterised in that, said at leasi one operating parameter is selected from a generator torque signal, a rotor torque signal, a vibration signal, or a blade bending moment signal.
    [9] 9. A method according to any one of claims 1 to 8, characterised in that, said method further comprises at least the step of: operating the wind turbine in a load protective mode, wherein the configuration of the wind turbine in said load protective mode is selected according tc the identified wind distribution pattern, or storing the wind distribution in a database, e.g. transmitting said stored wine distribution to a remote control system or monitoring unit.
    [10] 10. A wind turbine comprising: a rotor rotary arranged relative to a nacelle, the rotor comprises at least twe wind turbine blades mounted to a rotatable hub, wherein the at least two wine turbine blades define a rotor plane, the nacelle being arranged on top of a wind turbine tower, an angular sensor configured to measure a rotational position of the rotor, at least a second sensor configured to measure at least one operating parametei of the wind turbine, a control system electrically connected to the angular sensor and the at leas second sensor, characterised in that, the control system is configured to determine a wind distribution over the rotor plane, e.g. in at least a horizontal direction or a vertical direction, wherein the control system is further configurec to compare said wind distribution to a plurality of predetermined wind distribution patterns and to identify a match between the wind distribution and one of said plurality of predetermined wind distribution patterns.
    [11] 11. A wind turbine according to claim 10, characterised in that, said control systeir is configured to determine at least one wind turbine blade passing signal from said a least one operating parameter or said rotational position, and to calculate the characteristics, e.g. at least an amplitude or a phase, of said one wind turbine blade passing signal.
    [12] 12. A wind turbine according to claim 11, characterised in that, each of the predetermined wind distribution patterns is defined by a unique relationship between saic each wind distribution pattern and predetermined characteristics of the wind turbine blade passing signal, wherein the control system is configured to compare the calculated characteristic to said predetermined characteristics for identifying a match.
    [13] 13. A wind turbine according to any one of claims 10 to 12, characterised in that said control system is configured to operate the wind turbine in a normal operatior mode and in at least one load protective mode, wherein the control system in the a least one load protective mode is configured to apply a protective action to the wine turbine based on the identified wind distribution pattern.
    类似技术:
    公开号 | 公开日 | 专利标题
    EP3384154B1|2020-10-28|A wind turbine and a method of operating a wind turbine for reducing edgewise vibrations
    US9004862B2|2015-04-14|Calibration of wind turbine sensor
    US7895018B2|2011-02-22|Event monitoring via combination of signals
    US10907615B2|2021-02-02|Method of correcting rotor imbalance and wind turbine thereof
    EP3344870B1|2021-04-21|A wind turbine and a method of operating a wind turbine with a rotational speed exclusion zone
    CN106795857B|2020-04-10|Method of improving the balance of a rotor in a wind turbine and wind turbine system
    US10907617B2|2021-02-02|Diagnostic system and method for use in a wind turbine
    KR20130028923A|2013-03-20|Control device for a wind power plant
    US20170335829A1|2017-11-23|Method of Identifying a Wind Distribution Pattern Over the Rotor Plane and a Wind Turbine Thereof
    EP2678556B1|2018-04-25|A safety system for a wind turbine
    US10774810B2|2020-09-15|System and method for estimating high bandwidth tower deflection for wind turbines
    WO2018233787A1|2018-12-27|A method for determining wind turbine blade edgewise load recurrence
    EP2881581B1|2018-01-10|Wind turbine control system and method
    EP3237753B1|2019-02-20|A method of operating a wind turbine
    TWI729349B|2021-06-01|Wind power generation device and wind power generation system
    WO2021156287A1|2021-08-12|Wake monitoring, wake management and sensory arrangements to such
    CN113252938A|2021-08-13|Method for determining the wind speed in the rotor plane of a wind turbine
    同族专利:
    公开号 | 公开日
    CN107420269A|2017-12-01|
    EP3249217A1|2017-11-29|
    DK179333B1|2018-05-07|
    US20170335829A1|2017-11-23|
    CN107420269B|2019-12-13|
    引用文献:
    公开号 | 申请日 | 公开日 | 申请人 | 专利标题
    US20100014969A1|2006-10-02|2010-01-21|Kitchener Clark Wilson|Wind turbine with blade pitch control to compensate for wind shear and wind misalignment|
    EP2175131A2|2008-10-10|2010-04-14|General Electric Company|Apparatus and method for continuous pitching of wind turbine blades|
    EP2317327A1|2009-10-28|2011-05-04|SSB Wind Systems GmbH & Co. KG|Wind sensor system using blade signals|
    DE102013004446A1|2013-03-15|2014-09-18|Abb Ag|Method for estimating wind speed distribution profiles in wind turbines|
    ES2655286T3|2007-03-16|2018-02-19|Vestas Wind Systems A/S|Method for monitoring the condition of a rotor of a wind power plant|
    US7950901B2|2007-08-13|2011-05-31|General Electric Company|System and method for loads reduction in a horizontal-axis wind turbine using upwind information|
    NL2001878C2|2008-08-07|2010-02-09|Stichting Energie|System and method for compensating rotor imbalance for a wind turbine.|
    CN101684774B|2008-09-28|2012-12-26|通用电气公司|Wind power generation system and wind measuring method of wind power generator|
    DE102009030886A1|2009-06-29|2010-12-30|Robert Bosch Gmbh|Wind turbine with a variety of wind energy devices and methods for controlling the wind turbine|
    US7933744B2|2009-08-28|2011-04-26|General Electric Company|System and method for managing wind turbines and enhanced diagnostics|
    US20110054825A1|2009-08-28|2011-03-03|General Electric Company|System and method for managing wind turbines|
    KR101634846B1|2010-04-19|2016-06-29|보벤 프로퍼티즈 게엠베하|Method for the operation of a wind turbine|
    EP2421130A1|2010-08-20|2012-02-22|Siemens Aktiengesellschaft|Method and sensor unit for determining the current usage level indicator value of an electric and/or electronic component in a wind turbine|
    GB201109897D0|2011-06-13|2011-07-27|Romax Technology Ltd|A method for managing wind farms|
    CN102506016A|2011-10-27|2012-06-20|镇江明润信息科技有限公司|Wind power station equipment status inspection device|
    CN103225588B|2012-01-31|2015-05-13|北京能高自动化技术股份有限公司|Wind power generation power curve optimization method based on pattern recognition technology|
    JP6151030B2|2012-02-02|2017-06-21|三菱重工業株式会社|Wind power generator and operation control method thereof|
    EP2859223B1|2012-06-06|2018-05-23|Vestas Wind Systems A/S|A wind turbine with a load controller|
    WO2013182204A1|2012-06-08|2013-12-12|Vestas Wind Systems A/S|A method of operating a wind turbine as well as a system suitable therefore|
    CN104454350B|2013-09-23|2019-01-08|通用电气公司|Wind turbine and its control method for reducing rotor unbalance load|
    ES2537476B1|2013-12-04|2016-03-23|Acciona Windpower, S.A.|Wind turbine control system and method|
    CN103899498A|2014-04-16|2014-07-02|左浩俊|Data collecting and processing system of wind turbine generator|
    KR101571325B1|2014-05-09|2015-11-24|삼성중공업 주식회사|Method of estimating yaw error and apparatus thereof|
    US9995276B2|2014-06-19|2018-06-12|Vestas Wind Systems A/S|Control of wind turbines in response to wind shear|
    CN104763585B|2015-02-06|2017-12-22|湘潭大学|Wind turbines dynamic reconfiguration method based on distributed data collection|
    CN204941779U|2015-09-02|2016-01-06|上海僖舜莱机电设备制造有限公司|Vertical axis aerogenerator and wind wheel arrangements for speed regulation thereof|EP3499022A1|2017-12-12|2019-06-19|General Electric Company|Methods of operating a wind turbine|
    DE102018130636A1|2018-12-03|2020-06-04|Wobben Properties Gmbh|Method for operating a wind turbine|
    DK201970353A1|2019-05-29|2021-02-09|Envision Energy Jiangsu Co Ltd|A wind turbine and a method of operating the wind turbine|
    法律状态:
    优先权:
    申请号 | 申请日 | 专利标题
    CN201610345279.6A|CN107420269B|2016-05-23|2016-05-23|method for identifying a wind distribution pattern on a rotor plane and wind turbine implementing the method|
    [返回顶部]