JP7438288B2 - How to diagnose wind turbine blades - Google Patents

Description

The present disclosure relates to a method for diagnosing a wind turbine blade.

It is known to detect damage to wind turbine blades by measuring strain in the wind turbine blades of a wind power generator (for example, see Patent Document 1).

Japanese Patent Application Publication No. 2016-156674

In the device described in Patent Document 1, distortion of the wind turbine blade can be measured in real time using a sensor attached to the wind turbine blade. Therefore, it is possible to grasp the load acting on the wind turbine blades during operation of the wind power generator.
Therefore, for example, a device installed on the wind turbine blade can be used to manage the operation of the wind power generator, such as by detecting that an abnormal load is applied to the wind turbine blade while the wind power generator is in operation, and stopping the operation of the wind power generator. It is also possible to use sensors.

However, if the wind turbine blades of an existing wind power generation device are not provided with the above-mentioned sensors, operation management using the sensors is not possible.

In view of the above-mentioned circumstances, at least one embodiment of the present disclosure aims to provide a method for diagnosing a wind turbine blade by measuring strain of the wind turbine blade in real time in an existing wind power generation device.

(1) A method for diagnosing a wind turbine blade according to at least one embodiment of the present disclosure includes:
A method for diagnosing wind turbine blades of an existing wind power generation device, the method comprising:
a step of additionally installing a sensor for detecting distortion of the wind turbine blade;
acquiring a detection signal from the sensor;
calculating a load acting on the wind turbine blade based on the detection signal;
Equipped with
In the step of additionally installing the sensor, a first sensor is installed at one of two positions facing each other along the cord direction of the wind turbine blade among circumferential positions in the blade root part centered on the blade axis of the wind turbine blade. and a second sensor is installed at one of two positions facing each other along a direction perpendicular to the code direction,
In the step of calculating the load, the load is calculated based on a detection signal from the first sensor and a detection signal from the second sensor.

(2) A method for diagnosing a wind turbine blade according to at least one embodiment of the present disclosure includes:
A method for diagnosing wind turbine blades of an existing wind power generation device,
a step of additionally installing a sensor for detecting distortion of the wind turbine blade;
acquiring a detection signal from the sensor;
calculating a load acting on the wind turbine blade based on the detection signal;
Detecting the presence or absence of an abnormality in the wind turbine blade based on the detection signal;
Equipped with.

According to at least one embodiment of the present disclosure, it is possible to provide a method for diagnosing wind turbine blades by measuring strain in the wind turbine blades in real time in an existing wind power generation device.

1 is a schematic diagram showing the overall configuration of a wind turbine of a wind power generation device according to some embodiments. FIG. 2 is a schematic diagram showing a cross section perpendicular to the longitudinal direction of the wind turbine blade at a blade root portion of the wind turbine blade according to an embodiment. FIG. 7 is a schematic diagram showing a cross section perpendicular to the longitudinal direction of the wind turbine blade at a blade root portion of a wind turbine blade according to another embodiment. FIG. 2 is a block diagram illustrating the configuration of an embodiment regarding processing of a converted signal. FIG. 3 is a block diagram regarding the configuration of another embodiment regarding processing of converted signals. FIG. 3B is a perspective view schematically showing the inside of the hub in the embodiment shown in FIG. 3B. 3 is a flowchart showing a processing procedure of a method for diagnosing a wind turbine blade according to some embodiments. FIG. 3 is a block diagram related to calculation of a leeward bending moment per blade and a torque component per blade. 7 is a more detailed block diagram of a portion of the block diagram of FIG. 6; FIG. FIG. 3 is a block diagram regarding calculation of a composite torque component. FIG. 2 is a block diagram for calculation of bending moments for the wind turbine rotor and bending moments for the nacelle. It is a figure for explaining the bending moment about a wind turbine blade. FIG. 3 is a diagram for explaining a bending moment about a wind turbine rotor. It is a figure for explaining the bending moment about a nacelle. It is a flowchart which shows the processing procedure regarding the calibration process of a sensor. 1 is an example of a schematic cross-sectional view of a wind turbine blade according to some embodiments in a cross section perpendicular to a blade axis; FIG.

Hereinafter, some embodiments of the present disclosure will be described with reference to the accompanying drawings. However, the dimensions, materials, shapes, relative arrangements, etc. of the components described as embodiments or shown in the drawings are not intended to limit the scope of the present disclosure, and are merely illustrative examples. do not have.
For example, expressions expressing relative or absolute positioning such as "in a certain direction,""along a certain direction,""parallel,""orthogonal,""centered,""concentric," or "coaxial" are strictly In addition to representing such an arrangement, it also represents a state in which they are relatively displaced with a tolerance or an angle or distance that allows the same function to be obtained.
For example, expressions such as "same,""equal," and "homogeneous" that indicate that things are in an equal state do not only mean that things are exactly equal, but also have tolerances or differences in the degree to which the same function can be obtained. It also represents the existing state.
For example, expressions expressing shapes such as squares and cylinders do not only refer to shapes such as squares and cylinders in a strict geometric sense, but also include uneven parts and chamfers to the extent that the same effect can be obtained. Shapes including parts, etc. shall also be expressed.
On the other hand, the expressions "comprising,""comprising,""comprising,""containing," or "having" one component are not exclusive expressions that exclude the presence of other components.

(overall structure)
First, the configuration of a wind power generation device to which a wind turbine blade diagnosis method according to some embodiments is applied will be described. FIG. 1 is a schematic diagram showing the overall configuration of a wind turbine of a wind power generator according to some embodiments. As shown in the figure, the windmill 1a includes a windmill rotor 5 including at least one windmill blade 3 and a hub 2 to which the windmill blade 3 is attached. The wind turbine rotor 5 is provided at the top of the tower 6, and is rotatably supported by a nacelle 4 supported by the tower 6, so that when the wind turbine blades 3 receive wind, the wind turbine rotor 5 including the wind turbine blades 3 rotates. It has become.

In some embodiments, the wind turbine 1a is part of the wind power generator 1. The nacelle 4 accommodates a generator (not shown) and a power transmission mechanism (not shown) for transmitting rotation of the wind turbine rotor 5 to the generator. The wind power generator 1 is configured such that rotational energy transmitted from the wind turbine rotor 5 to the generator via the power transmission mechanism is converted into electrical energy by the generator.

FIG. 2A is a schematic diagram showing a cross section perpendicular to the longitudinal direction of the wind turbine blade at the blade root portion of the wind turbine blade according to one embodiment.
FIG. 2B is a schematic diagram showing a cross section of a blade root portion of a wind turbine blade according to another embodiment, which is perpendicular to the longitudinal direction of the wind turbine blade.
As shown in FIGS. 1 and 2A, the wind turbine 1a has four optical fiber sensors 12 (12A to 12D) attached to each blade root 3a of the wind turbine blade 3 as sensors for detecting distortion of the wind turbine blade 3. ). As shown in FIGS. 1 and 2A, the windmill 1a includes a light source/signal processing unit 10 including a light incidence section 17 and a light detection section 18.
As shown in FIG. 2B, the wind turbine 1a includes two optical fiber sensors 12 (12A and 12D) attached to each blade root 3a of the wind turbine blade 3 as sensors for detecting distortion of the wind turbine blade 3. You can leave it there. As shown in FIG. 2B, the windmill 1a may include a light source/signal processing unit 10 including a light incidence section 17 and a light detection section 18.

The optical fiber sensor 12 has a diffraction grating portion whose refractive index changes periodically in the longitudinal direction.
The optical fiber sensor 12 according to some embodiments is an FBG (Fiber Bragg Grating) sensor.
When light with a broadband spectrum is incident on the optical fiber sensor 12, the reflection at the diffraction grating part whose refractive index changes periodically will cause the light to be reflected from each other only for specific wavelengths depending on the grating interval (grating period). Interfering in a mutually reinforcing direction. As a result, the optical fiber sensor 12 reflects only specific wavelength components of the configuration and transmits other wavelengths.
When the strain applied to the optical fiber sensor 12 or the ambient temperature changes, the refractive index and grating spacing of the diffraction grating section change, and the wavelength λ _o of the reflected light changes in accordance with this change.
The difference between the wavelength λ _o of the reflected light in the optical fiber sensor 12 and the initial wavelength λ _i is affected by the strain ε _z and the temperature ΔT of the optical fiber sensor 12 . Therefore, if the initial wavelength λ _i is a predetermined value, the strain occurring in the optical fiber sensor 12 and the temperature of the optical fiber can be detected by measuring the wavelength λ _o of the reflected light.

The optical fiber sensor 12 and the light source/signal processing unit 10 are added later to the existing wind power generation device 1, as will be described later.
Four optical fiber sensors 12A to 12D are attached to the blade root portion 3a of each wind turbine blade 3 shown in FIG. 2A. As shown in FIG. 2A, the blade root portion 3a of the wind turbine blade 3 has a substantially circular cross section, and the optical fiber sensors 12A to 12D are attached to the wall surface of the blade root portion 3a at 90° intervals. In each wind turbine blade 3 shown in FIG. 2A, the four optical fiber sensors 12A, 12B, 12C, and 12D are arranged in this order on the ventral side (HP side) 21, trailing edge side 24, and dorsal side (LP side) of the wind turbine blade 3, respectively. side) 22 and the leading edge side 23 on the wall surface of the blade root portion 3a. More specifically, in each wind turbine blade 3 shown in FIG. A leading edge side sensor (optical fiber sensor 12D) and a trailing edge side sensor (optical fiber sensor 12B) are arranged at two positions on the opposing leading edge side 23 and trailing edge side 24, respectively, and along the direction perpendicular to the code C direction. A ventral side sensor (optical fiber sensor 12A) and a dorsal side sensor (optical fiber sensor 12C) are respectively disposed at two positions on the ventral side 21 and the dorsal side 22 which face each other.

In each of the wind turbine blades 3 shown in FIG. 2A, when strain occurs in the blade root portion 3a, strain occurs in each of the optical fiber sensors 12A to 12D in accordance with the strain at each attached location.
Further, these four optical fiber sensors 12A, 12B, 12C, and 12D are connected in series in this order by an optical fiber cable 16. Connectors (13, 15) are provided at both ends of the optical fiber cable 16, and connect the optical fiber cable 14 and the optical fiber sensors 12A to 12D to connect the optical fiber sensors 12A to 12D and the light source/signal processing unit 10. An optical fiber cable 16 that connects the two in series is connected via the connector 13 or the connector 15.

Note that in each wind turbine blade 3 shown in FIG. 2A, four optical fiber sensors 12A to 12D are attached to one wind turbine blade 3, but in each wind turbine blade 3 shown in FIG. Two optical fiber sensors 12 are attached to the sensor. In each wind turbine blade 3 shown in FIG. 2B, the optical fiber sensor 12 has a leading edge side 23 facing along the code C direction of the wind turbine blade 3 among circumferential positions in the blade root portion 3a centered on the blade axis L0. Optical fiber sensor 12 disposed at one of two positions with the rear edge side 24, and one of two positions with the ventral side 21 and dorsal side 22 facing each other along the direction orthogonal to the code C direction. It is preferable to include an optical fiber sensor 12 disposed at. The two optical fiber sensors 12 attached to each wind turbine blade 3 shown in FIG. An optical fiber sensor 12D placed on the edge side 23, and an optical fiber sensor 12A placed on the ventral side 21 of two positions, the ventral side 21 and the dorsal side 22 facing each other along the direction orthogonal to the code C direction. That is.
Note that one of the two optical fiber sensors 12 attached to each wind turbine blade 3 shown in FIG. Instead of the optical fiber sensor 12D placed on the leading edge side 23 of the two positions, the optical fiber sensor 12B may be placed on the trailing edge side 24. In addition, the other one of the two optical fiber sensors 12 attached to each wind turbine blade 3 shown in FIG. The optical fiber sensor 12C may be placed on the dorsal side 22 instead of the optical fiber sensor 12A placed on the ventral side 21 of the two positions.

Further, in the above description, an example has been described in which the optical fiber sensor 12 is attached to the blade root portion 3a of the wind turbine blade 3, but the position where the optical fiber sensor 12 is attached to the wind turbine blade 3 is not limited to the blade root portion 3a. For example, the optical fiber sensor 12 may be attached to a hub structure (not shown) that supports the wind turbine blade 3 so that the pitch angle can be changed.

Further, for example, the optical fiber sensor 12 may be attached to a blade root bearing, that is, a pitch bearing 300 shown in FIG. 4, which will be described later. Specifically, the optical fiber sensor 12 may be attached to a fastening member such as a bolt or nut (not shown), which will be described below.
For example, in the wind turbine 1a according to some embodiments, the pitch bearing 300 has a configuration in which rolling elements (not shown) are held between an inner ring 304 and an outer ring 306 that are rotatable relative to each other. The inner ring 304 is fixed to the wind turbine blade 3 with unillustrated bolts, and the outer ring 306 is fixed to the hub 2 with unillustrated bolts. By disposing such a pitch bearing 300 between the hub 2 and the wind turbine blade 3, the wind turbine blade 3 fixed to the inner ring 304 can move in the pitch direction (wind turbine blade 3) with respect to the hub 2 fixed to the outer ring 306. It is configured to be able to rotate around the wing axis L0 of No. 3).

In some embodiments, the optical fiber sensor 12 may be attached to the aforementioned bolt (not shown) or a fastening member such as a nut used with the bolt.

Further, in the above description, the optical fiber sensor 12 has been described as an example of a sensor for detecting the distortion of the wind turbine blade 3, but the sensor for detecting the distortion of the wind turbine blade 3 is limited to the optical fiber sensor 12. Not done. The sensor for detecting strain in the wind turbine blade 3 may be, for example, a metal resistance strain gauge or a semiconductor strain gauge.
Note that if the sensor for detecting the strain of the wind turbine blade 3 is a metal resistance type strain gauge or a semiconductor strain gauge, for example, a strain gauge signal conditioner 11 (see FIG. 3B) may be used instead of the light source/signal processing unit 10. use

The light source/signal processing unit 10 including the light incidence section 17 and the light detection section 18 is installed inside the hub 2 in some embodiments.
The light input section 17 is configured to input light to the optical fiber sensors 12A to 12D. The light incidence section 17 includes, for example, a light source capable of emitting light with a broadband spectrum.
Furthermore, the light detection section 18 is configured to detect the reflected light from the diffraction grating section of each of the optical fiber sensors 12A to 12D.

Broadband light emitted from the light incidence section 17 propagates through the optical fiber cables (14, 16) and is incident on each of the optical fiber sensors 12A to 12D. In each of the optical fiber sensors 12A to 12D, among the incident light, only light having a specific wavelength depending on the strain and temperature of each diffraction grating is reflected by the diffraction grating. The reflected light reflected by the diffraction grating section propagates through the optical fiber cable (16, 14) and is detected by the light detection section 18.
Note that the optical fiber sensors 12A to 12D are connected in series, and the distances from the light incidence section 17 to each of the optical fiber sensors 12A to 12D and the distances from the optical fiber sensors 12A to 12D to the light detection section 18 are different from each other. different. Therefore, in the light detection section 18, detection is performed based on the time elapsed from when the light is emitted by the light incidence section 17 until the reflected light from each of the optical fiber sensors 12A to 12D is detected by the light detection section 18. It is possible to determine which optical fiber sensor 12 each reflected light comes from.

In the following description, the reflected light from the diffraction grating portion of each of the optical fiber sensors 12A to 12D is also referred to as a detection signal, and the optical signal from the optical fiber sensors 12A to 12D is also referred to as a detection signal.
Note that if the sensor for detecting the strain of the wind turbine blade 3 is, for example, a metal resistance type strain gauge or a semiconductor strain gauge, the signal from the metal resistance type strain gauge or semiconductor strain gauge is also referred to as a detection signal.
Further, in the following description, the electric signal related to the distortion of the wind turbine blade 3 output from the light source/signal processing unit 10 is also referred to as a conversion signal of a detection signal or simply a conversion signal.
Similarly, if the sensor for detecting the strain on the wind turbine blade 3 is, for example, a metal resistance type strain gauge or a semiconductor strain gauge, the electric signal related to the strain on the wind turbine blade 3 output from the strain gauge signal conditioner 11 is It is also referred to as a conversion signal of a detection signal, or simply a conversion signal.

Furthermore, in the following description, the wind turbine rotor 5 of the wind turbine 1a will be described as having, for example, three wind turbine blades 3. When it is necessary to distinguish between the three wind turbine blades 3, one of the three wind turbine blades 3 is referred to as the first wind turbine blade 31, and two wind turbine blades other than the first wind turbine blade 31 are referred to as the first wind turbine blade 31. One of the wind turbine blades 3 is referred to as a second wind turbine blade 32, and the remaining wind turbine blades 3 are referred to as a third wind turbine blade 33.

In the following description, the optical fiber sensors 12 (12A to 12D) are also simply referred to as sensors 12 (12A to 12D) in the present specification and drawings.

FIG. 3A is a block diagram illustrating a configuration of an embodiment related to processing a transformed signal.
The embodiment shown in FIG. 3A is a block diagram in a case where the sensor for detecting distortion of the wind turbine blade 3 is an optical fiber sensor 12.
Moreover, the embodiment shown in FIG. 3A is also an embodiment that can be implemented mainly when the existing wind power generation device 1 is a wind power generation device manufactured by the company.
In the embodiment shown in FIG. 3A, the existing wind power generator 1 includes a hub PLC 51 disposed within the hub 2 and a nacelle PLC 53 disposed within the nacelle 4 as the existing arithmetic unit 50 included in the existing wind power generation device 1.
In the embodiment shown in FIG. 3A, as described above, the optical fiber sensor 12 and the light source/signal processing unit 10 are added later to the existing wind power generator 1. In the embodiment shown in FIG. 3A, a light source/signal processing unit 10 is provided for every three wind turbine blades 3.

In the embodiment shown in FIG. 3A, the light source/signal processing unit 10 is configured to be supplied with power from a hub power supply 57, which is an existing power supply in the hub 2.

In the embodiment shown in FIG. 3A, the converted signal from the light source and signal processing unit 10 is output to the hub PLC 51.
In the embodiment shown in FIG. 3A, based on the conversion signal from the light source/signal processing unit 10, the hub PLC 51 calculates the bending moment, etc. that acts on the wind turbine blade 3 due to the wind load, as will be described in detail later. You can also do this.

Note that when the hub PLC 51 calculates the bending moment and the like acting on the wind turbine blade 3 due to wind load, the hub PLC 51 outputs the calculation result to the nacelle PLC 53.
In this case, the nacelle PLC 53 outputs the calculation result received from the hub PLC 51 to the SCADA 61, which is a device for remotely monitoring the wind power generator 1.

Further, in the embodiment shown in FIG. 3A, the bending moment and the like acting on the wind turbine blade 3 due to wind load may be calculated in the nacelle PLC 53 based on the conversion signal from the light source/signal processing unit 10. In this case, the hub PLC 51 outputs the converted signal from the light source/signal processing unit 10 to the nacelle PLC 53.
The nacelle PLC 53 calculates the bending moment, etc. that acts on the wind turbine blade 3 due to the wind load, based on the conversion signal from the light source/signal processing unit 10 received via the hub PLC 51 . Then, the nacelle PLC 53 outputs the calculation result to the SCADA 61.

FIG. 3B is a block diagram regarding the configuration of another embodiment regarding processing of a converted signal.
The embodiment shown in FIG. 3B is a block diagram in a case where the sensor for detecting strain in the wind turbine blade 3 is a sensor 42 such as a metal resistance strain gauge or a semiconductor strain gauge.
Further, the embodiment shown in FIG. 3B mainly applies when the existing wind power generation device 1 is a wind power generation device manufactured by another company, or even if the existing wind power generation device 1 is a wind power generation device manufactured by the company itself. This is also an embodiment that can be implemented when there is a desire to calculate the bending moment, etc. that acts on the wind turbine blade 3 due to wind load with a calculation device having a higher processing capacity than the existing calculation device 50.
The embodiment shown in FIG. 3B includes a wind turbine PLC 55 that is an existing calculation device 50 included in the existing wind power generation device 1.
Note that the wind turbine PLC 55 may be the hub PLC 51 described above, the nacelle PLC 53, or the hub PLC 51 and the nacelle PLC 53.

In the embodiment shown in FIG. 3B, the sensor 42 and the strain gauge signal conditioner 11 are added to the existing wind power generator 1 later.
Further, in the embodiment shown in FIG. 3B, power is mainly supplied to the calculation device (additional calculation device) 63 for calculating the bending moment etc. that acts on the wind turbine blade 3 due to wind load, and the additional calculation device 63. A possible power storage device 65 is added to the existing wind power generation device 1 later.
In the embodiment shown in FIG. 3B, a communication device 67 and a relay 69 are added to the existing wind power generation device 1 later.

In the embodiment shown in FIG. 3B, the additional calculation device 63 may be, for example, a small personal computer, a so-called single board computer, or the like.
FIG. 4 is a perspective view schematically showing the inside of the hub 2 in the embodiment shown in FIG. 3B. In the embodiment shown in FIG. 3B, as shown in FIG. 4A, the additional calculation device 63 may be arranged, for example, near the existing calculation device 50 of the hub 2.

Note that the computing power of a small personal computer, a so-called single board computer, etc. that is additionally installed as the additional computing device 63 is considered to be higher than that of the existing computing device 50 provided in the existing wind power generation device 1. Therefore, as in the embodiment shown in FIG. 3B, for example, when a small personal computer or a so-called single board computer is additionally installed as the additional calculation device 63, there is a case where the additional calculation device 63 is not additionally installed. The sampling frequency of the sensor 12 may be set to a higher frequency than the above.

In the embodiment shown in FIG. 3B, the power storage device 65 is a device that can supply power to the strain gauge signal conditioner 11 and the additional calculation device 63 while receiving power from the hub power supply 57, for example. This device has a power storage function that can supply power to the strain gauge signal conditioner 11 and the additional calculation device 63 for a specified period of time even if the power supply from the hub power supply 57 is cut off.
For example, the power storage device 65 may include a rechargeable and dischargeable secondary battery. Note that the power storage device 65 may include a primary battery instead of a secondary battery.

In the embodiment shown in FIG. 3B, the communication device 67 wirelessly transmits calculation results such as the bending moment acting on the wind turbine blade 3 due to the wind load calculated by the additional calculation device 63 to a device external to the wind power generation device 1. It is a device for transmitting data.
In the embodiment shown in FIG. 3B, the relay 69 is used, for example, to output an alarm signal or the like to the wind turbine PLC 55 based on the calculation results such as the bending moment calculated by the additional calculation device 63. Note that in the embodiment shown in FIG. 3B, the relay 69 is not necessarily required, and for example, calculations in the additional calculation device 63 may be output directly to the wind turbine PLC 55.

In the embodiment shown in FIG. 3B, the converted signal from the strain gauge signal conditioner 11 is output to an additional computing device 63.
In the embodiment shown in FIG. 3B, the additional calculation device 63 is configured to calculate the bending moment, etc. that acts on the wind turbine blade 3 due to the wind load, based on the conversion signal from the strain gauge signal conditioner 11. .
In the embodiment shown in FIG. 3B, the additional calculation device 63 can wirelessly transmit the calculated results of the bending moment and the like to a device external to the wind power generation device 1 via the communication device 67.
Further, in the embodiment shown in FIG. 3B, the additional calculation device 63 determines whether or not to output an alarm signal etc. to the wind turbine PLC 55 based on the calculation result of the calculated bending moment, etc. If it is determined that a signal etc. should be outputted, an alarm signal etc. can be outputted to the wind turbine PLC 55 via the relay 69.

In the embodiment shown in FIG. 3B, even if some malfunction occurs and the hub power supply 57 is out of power, the power storage device 65 supplies power to the strain gauge signal conditioner 11 and the additional calculation device 63 for a specified period of time. can be supplied. That is, even if the power supply from the wind power generator 1 side is cut off due to some kind of malfunction, the additional calculation device 63 can be operated with the power from the power storage device 65.

(Diagnosis method for wind turbine blades)
Hereinafter, methods for diagnosing wind turbine blades according to some embodiments will be described.
FIG. 5 is a flowchart illustrating a processing procedure of a method for diagnosing a wind turbine blade according to some embodiments.
The wind turbine blade diagnosing method according to some embodiments is a diagnosing method of the wind turbine blade 3 of the existing wind power generation device 1, and includes an additional installation step S10, a detection signal acquisition step S30, a load calculation step S50, Equipped with Note that the wind turbine blade diagnosis method according to some embodiments may further include an external transmission step S70 and an abnormality detection step S90.

(Additional installation process S10)
In the wind turbine blade diagnosis method according to some embodiments, the additional installation step S10 is a step of additionally installing, for example, an optical fiber sensor 12 as a sensor for detecting distortion of the wind turbine blade 3 in the existing wind power generation device 1. It is. In the additional installation step S10 according to some embodiments, the sensor for detecting the strain of the wind turbine blade 3 may be, for example, a sensor 42 such as a metal resistance strain gauge or a semiconductor strain gauge.
For example, as shown in FIGS. 1 and 2A, the sensor to be additionally installed in the additional installation step S10 according to some embodiments may be placed in the code C direction among the circumferential positions in the blade root 3a centered on the blade axis L0. It is preferable to include a leading edge side sensor (optical fiber sensor 12D) and a trailing edge side sensor (optical fiber sensor 12B) disposed at two positions, respectively, on the leading edge side 23 and the trailing edge side 24 facing each other along. The sensor additionally installed in the additional installation step S10 according to some embodiments has a ventral side 21 and a dorsal side 22 facing each other along the direction orthogonal to the code C direction, as shown in FIGS. 1 and 2A, for example. It is preferable to include a ventral side sensor (optical fiber sensor 12A) and a dorsal side sensor (optical fiber sensor 12C) arranged at two positions, respectively. Note that instead of the optical fiber sensors 12A to 12D, a sensor 42 such as a strain gauge or a semiconductor strain gauge may be installed.

In the additional installation step S10 according to some embodiments, two positions on the leading edge side 23 and the trailing edge side 24 that face each other along the code C direction among the circumferential positions in the blade root 3a centering on the blade axis L0 are For example, an optical fiber sensor 12 is disposed as a first sensor at one of the two positions, and a second sensor is disposed at one of the two positions of the ventral side 21 and the dorsal side 22 facing each other along the direction orthogonal to the code C direction. For example, an optical fiber sensor 12 may be installed. Note that instead of the optical fiber sensor 12, a sensor 42 such as a strain gauge or a semiconductor strain gauge may be installed.

In the additional installation step S10 according to some embodiments, the additionally installed sensors 12 and 42 include an abnormality detection sensor that detects distortion of the wind turbine blade 3 in order to detect the presence or absence of an abnormality in the wind turbine blade 3. Good. These abnormality detection sensors (sensors 12, 42) may also be used as sensors 12, 42 that output detection signals for calculating the load acting on the wind turbine blades 3, as will be described later. It may be a dedicated sensor 12, 42 for detecting the presence or absence of.

If the abnormality detection sensor is also used as the sensor 12, 42 that outputs a detection signal for calculating the load acting on the wind turbine blade 3, the number of additional sensors 12, 42 can be suppressed, and the number of sensors 12, 42 can be reduced. Additional installation costs can be reduced.

If the abnormality detection sensor is a dedicated sensor 12 or 42 for detecting the presence or absence of an abnormality in the wind turbine blade 3, for example, the frequency of execution of arithmetic processing for detecting the presence or absence of an abnormality in the wind turbine blade 3 may be suppressed. As a result, the processing of the detection signal from the abnormality detection sensor can be simplified, and the calculation load on the existing calculation device 50 and additional calculation device 63 provided in the existing wind power generation device 1, such as the hub PLC 51, can be suppressed.

Further, in the additional installation step S10 according to some embodiments, as shown in FIGS. 3A and 3B, in the existing wind power generation device 1, the light source/signal processing unit 10, strain gauge signal conditioner 11, etc. A device is additionally installed to receive a detection signal from a sensor for detecting the distortion of the wind turbine blade 3 and output a conversion signal that is an electrical signal related to the distortion of the wind turbine blade 3.

In the additional installation step S10 according to some embodiments, as shown in FIG. 3B, an additional calculation device 63 may be additionally installed in the existing wind power generation device 1, and a power storage device 65 may be additionally installed. Often, a communication device 67 or a relay 69 may be additionally provided.

In addition, in the additional installation step S10 according to some embodiments, a worker adds each device to be added as described above to the existing wind power generation device 1.

(Detection signal acquisition step S30)
In the wind turbine blade diagnosis method according to some embodiments, the detection signal acquisition step S30 is a step of acquiring a detection signal from a sensor (for example, the optical fiber sensor 12 or the sensor 42) for detecting distortion of the wind turbine blade 3. It is.
In the detection signal acquisition step S30 according to some embodiments, a detection signal from a sensor for detecting strain in the wind turbine blade 3 is acquired by the light source/signal processing unit 10 or the strain gauge signal conditioner 11. Then, in the detection signal acquisition step S30 according to some embodiments, the converted signals from the light source/signal processing unit 10 and the strain gauge signal conditioner 11 are used in the hub PLC 51, etc. It is output to the device 50 and the additional calculation device 63.

Note that in the detection signal acquisition step S30), the sampling frequency for acquiring the detection signal may be changed based on the rotational speed of the wind turbine blade 3.
If the sampling frequency for acquiring detection signals is constant regardless of the rotational speed of the windmill blades 3, the larger the rotational speed of the windmill blades 3, the greater the azimuth angle from acquisition of a detection signal to acquisition of the next detection signal. The amount of change increases. Therefore, as the rotational speed of the wind turbine blade 3 increases, the number of detection signals that can be obtained during one rotation of the wind turbine blade 3 decreases, and fluctuations in the bending moment acting on the wind turbine blade 3 due to wind load are detected. Accuracy will decrease.
If the sampling frequency for acquiring the detection signal is changed based on the rotational speed of the windmill blade 3, more specifically, if the sampling frequency for acquiring the detection signal is increased as the rotational speed of the windmill blade 3 increases, the windmill blade Even if the rotational speed of No. 3 increases, it is possible to suppress a decrease in the accuracy of detecting the above-mentioned fluctuations.

(Load calculation step S50)
In the wind turbine blade diagnosis method according to some embodiments, the load calculation step S50 is a step of calculating the load acting on the wind turbine blade based on the detection signal.
In the load calculation step S50 according to some embodiments, based on the detection signal from the sensor for detecting the strain of the wind turbine blade 3, more specifically, the light source/signal processing unit 10 and the strain gauge signal condition are Based on the conversion signal from the wind turbine blade 11, the bending moment, etc. that acts on the wind turbine blade 3 due to the wind load is calculated in the existing calculation device 50 included in the existing wind power generation device 1, such as the hub PLC 51, or in the additional calculation device 63. do.
In addition, in the load calculation step S50 according to some embodiments, coordinate transformation is performed based on the calculated bending moment acting on the wind turbine blade 3, etc., so that the bending moment acting on the wind turbine rotor 5 and the nacelle 4 are Calculate the acting bending moment.
The load calculation step S50 according to some embodiments may be performed by an existing calculation device 50 included in the existing wind power generation device 1, such as the hub PLC 51, or may be performed by an additional calculation device 63. The specific contents of the calculation process in the load calculation step S50 will be described in detail later.

Note that the calculation in the load calculation step S50 according to some embodiments may be performed by the existing calculation device 50 included in the existing wind power generation device 1, or may be performed by the additional calculation device 63.
When the calculation in the load calculation step S50 is performed by the existing calculation device 50 included in the existing wind power generation device 1, the number of additional calculation devices (additional calculation device 63) can be suppressed. In providing a method for diagnosing the wind turbine blades 3 in the power generation device 1 by measuring the strain of the wind turbine blades 3 in real time, the cost of adding an arithmetic device can be reduced.

Generally, the existing calculation device 50 provided in the existing wind power generation device 1 is not designed to calculate the load acting on the wind turbine blade 3 based on the detection signal from the sensor 12. The computing power may not be sufficient to compute the load compared to the computing capability of a small personal computer, a so-called single board computer, etc. that is additionally installed as the additional computing device 63.
If the additional calculation device 63 is added in the addition step S10 and the calculation in the load calculation step S50 is performed by the additional calculation device 63, the additional calculation device 63 has sufficient calculation capacity to calculate the load. The load can be calculated by adding a calculating device 63.

(External transmission step S70)
In the wind turbine blade diagnosis method according to some embodiments, the external transmission step S70 is a step of transmitting the load information calculated in the load calculation step S50 to the outside of the wind power generator.
In the external transmission step S70 according to some embodiments, for example, as shown in FIG. wirelessly to the device. Note that if the bending moment and the like are calculated in the existing calculation device 50 of the existing wind power generator 1 in the load calculation step S50, the existing calculation device 50 calculates the communication device 67 in the external transmission step S70. The information may also be transmitted wirelessly to a device external to the wind power generator 1 via the wind power generator 1.

If the information on the load calculated in the step of calculating the load (load calculation step S50) is stored in a storage device etc. arranged inside the wind turbine generator 1, for example, if the information on the load is stored in the storage device etc. If an accident such as a collapse occurs and the storage device is destroyed, the information cannot be retrieved from the storage device, and the information cannot be used to investigate the cause of the accident.
By performing the external transmission step S70, if the above information is stored in a storage device etc. external to the wind power generator 1, even if an accident like the one described above occurs, the above information can be used to investigate the cause of the accident. can be used. That is, by performing the external transmission step S70, the possibility of the information being lost can be reduced.

(Abnormality detection step S90)
In the wind turbine blade diagnosing method according to some embodiments, the abnormality detection step S90 is a step of detecting the presence or absence of an abnormality in the wind turbine blade 3 based on the detection signal.
In the abnormality detection step S90 according to some embodiments, based on the detection signal from the sensor for detecting the distortion of the wind turbine blade 3, more specifically, the light source/signal processing unit 10 and the strain gauge signal condition are Based on the conversion signal from the wind turbine blade 11, the hub PLC 51 and the additional calculation device 63 determine whether or not there is an abnormality in the wind turbine blade 3.
The abnormality detection step S90 according to some embodiments may be performed by the existing calculation device 50 included in the existing wind power generation device 1, such as the hub PLC 51, or may be performed by the additional calculation device 63. The specific process of determining whether or not there is an abnormality in the wind turbine blade 3 in the abnormality detection step S90 will be described in detail later.
In the wind turbine blade diagnosing method according to some embodiments, by including the abnormality detection step S90, the presence or absence of an abnormality in the wind turbine blade 3 can be detected based on the distortion of the wind turbine blade 3.

(Calculation of downwind bending moment per blade and torque component per blade)
Hereinafter, among the specific calculation processing contents in the load calculation step S50 according to some embodiments, the calculation processing of the leeward bending moment per blade and the torque component per blade will be described.
FIG. 6 is a block diagram related to calculation of the leeward bending moment per blade and the torque component per blade. The calculation of the leeward bending moment per blade and the torque component per blade is performed by the existing calculation device 50 included in the existing wind power generation device 1, such as the hub PLC 51, or the additional calculation device 63. For convenience of explanation, in the following explanation, the leeward bending moment per blade and the torque component per blade are calculated in the additional calculation device 63. It may be implemented by the existing arithmetic unit 50 included in the wind power generation device 1.

Note that the block diagram in FIG. 6 is based on detection signals from two sensors (a first sensor and a second sensor) attached to each blade root 3a of the wind turbine blade 3, as shown in FIG. 2B, for example. This example is for calculating the leeward bending moment per blade and the torque component per blade. Note that the two sensors may be, for example, optical fiber sensors 12A and 12D, or may be two sensors 42 such as strain gauges or semiconductor strain gauges.
For example, as shown in FIG. 2A, the content of calculation based on detection signals from four sensors attached to each blade root 3a of the wind turbine blade 3 will be described later. Note that the four sensors may be, for example, optical fiber sensors 12A to 12D, or may be sensors 42 such as four strain gauges or semiconductor strain gauges.
In the following explanation, for convenience, the case will be mainly explained in which the sensor for detecting the strain on the wind turbine blade 3 is the optical fiber sensor 12. The same applies to the sensor 42 such as a gauge.

FIG. 7 shows the bending moment M _FLAP around the axis along the code C direction acting on the wind turbine blade 3 in the block diagram of FIG. FIG. 3 is a more detailed block diagram of the calculation of the bending moment M _EDGE about the axis.

First, explanation will be given based on FIG. 6.
The additional calculation device 63 corrects the strain εflap of the sensor 12A detected by the sensor 12A using the temperature correction unit 101f, thereby removing the influence of thermal elongation of the wind turbine blade 3 from the strain εflap.
The additional calculation device 63 eliminates the influence of centrifugal force and gravity acting on the wind turbine blade 3 by correcting the strain εflap excluding the influence of thermal elongation using the correction unit 102f.
The additional calculation device 63 corrects the strain εflap excluding the effects of centrifugal force and gravity using the correction unit 103f, thereby converting the strain εflap into a bending moment M _FLAP around the axis along the code C direction acting on the wind turbine blade 3. Calculate.

Note that the correction unit 103f calculates the bending moment M _FLAP using the following equation (1).
M _FLAP =k1×(εflap-b1)...(1)
Here, k1 is a strain moment calibration coefficient, which is a coefficient for converting strain into a bending moment.
b1 is the bending distortion correction coefficient ZERO adjustment, and is the distortion when the bending moment is zero.

Next, the pitch angle correction unit 104f decomposes the bending moment M _FLAP into a leeward bending moment component and a torque component of the wind turbine rotor 5 based on the pitch angle.

Similarly, the additional calculation device 63 corrects the strain εedge of the sensor 12D detected by the sensor 12D using the temperature correction unit 101e, thereby removing the influence of the thermal elongation of the wind turbine blade 3 from the strain εedge.
The additional calculation device 63 eliminates the influence of centrifugal force and gravity acting on the wind turbine blade 3 by correcting the strain εedge excluding the influence of thermal elongation using the correction unit 102e.
The additional calculation device 63 corrects the strain εedge excluding the effects of centrifugal force and gravity using the correction unit 103e, thereby converting the strain εedge into a code C acting on the wind turbine blade 3 and an axis perpendicular to the central axis of the wind turbine blade 3. Calculate the surrounding bending moment M _EDGE .

Note that the correction unit 103e calculates the bending moment M _EDGE using the following equation (2).
M _EDGE =k2×(εedge-b2)...(2)
Here, k2 is a distortion moment calibration coefficient, which is a coefficient for converting distortion into a bending moment.
b2 is the bending distortion correction coefficient ZERO adjustment, which is the distortion when the bending moment is zero.

Next, the pitch angle correction unit 104e decomposes the bending moment _MEDGE into a leeward bending moment component and a torque component of the wind turbine rotor 5 based on the pitch angle.

By adding together the component of the leeward bending moment found by the pitch angle corrector 104f and the component of the leeward bending moment found by the pitch angle corrector 104e, the leeward bending moment per blade is found. .

The torque component of the wind turbine rotor 5 obtained by the pitch angle correction section 104f and the torque component of the wind turbine rotor 5 obtained by the pitch angle correction section 104e are added together, and the cone angle correction section 105 performs correction based on the cone angle. By doing this, the torque component per blade can be determined.

Next, based on FIG. 7, bending moment M _FLAP acting on the wind turbine blade 3 around the axis along the code C direction, and around the axis perpendicular to the code C and the central axis (blade axis L0) of the wind turbine blade 3. The calculation of the bending moment M _EDGE will be explained.
The additional calculation device 63 uses a subtractor 201f to subtract the bending strain correction coefficient ZERO adjustment (b1) from the strain εflap of the sensor 12A detected by the sensor 12A, and subtracts the strain corresponding to the thermal elongation of the wind turbine blade 3 using a subtractor 202f. Subtract.

The additional calculation device 63 subtracts the distortion due to centrifugal force using a subtractor 203f. Specifically, by utilizing the fact that the square of the rotation speed of the wind turbine rotor 5 and the centrifugal force acting on the wind turbine blade 3 are directly proportional, the square of the rotation speed of the wind turbine rotor 5 and the centrifugal force acting on the wind turbine blade 3 are The relationship between the centrifugal force and the centrifugal force is determined in advance.
More specifically, by utilizing the fact that the square of the rotational speed of the windmill rotor 5 and the strain generated in the windmill blade 3 due to the centrifugal force acting on the windmill blade 3 are in a direct proportional relationship, the rotational speed of the windmill rotor 5 can be calculated. As a relationship between the square and the centrifugal force acting on the wind turbine blade 3, the tensile strain due to the centrifugal force per square of the rotational speed of the wind turbine rotor 5 of 1 rpm is determined in advance as a coefficient E. Then, using this coefficient, the additional calculation device 63 uses the multiplier 206 to calculate the tensile strain due to the centrifugal force.
Then, the additional calculation device 63 subtracts the tensile strain calculated by the multiplier 206 using a subtracter 203f.

Note that the coefficient E is determined as follows.
The centrifugal force acting on the wind turbine blade 3 determined from the detection signal from the sensor 12 at the azimuth angle where the blade axis L0 of the wind turbine blade 3 is closest to the vertical direction during the period before the existing wind power generation device 1 starts generating electricity. (More specifically, the relationship between the strain caused in the wind turbine blade 3 due to centrifugal force) and the square of the rotational speed of the wind turbine rotor 5 is obtained in advance. In other words, strain data at different rotational speeds of the wind turbine rotor 5 is acquired at a plurality of points. Then, the acquired data at multiple points are plotted on a graph in which the square of the rotational speed of the wind turbine rotor 5 is plotted on the horizontal axis, and the strain caused in the wind turbine blade 3 due to centrifugal force is plotted on the vertical axis. Let the slope of the obtained straight line be the above coefficient E.
This acquisition of the coefficient E may be performed by a worker, or may be performed periodically by the additional calculation device 63.

Note that the azimuth angle at which the blade axis L0 of the wind turbine blade 3 is closest in the vertical direction is the azimuth when the target wind turbine blade 3 is located at the 12 o'clock or 6 o'clock position when the wind turbine 1a is viewed from the front. It's about the corner.

By calculating as described above, it is possible to eliminate the influence of centrifugal force acting on the wind turbine blade 3 when determining the bending moment that acts on the wind turbine blade 3 due to the wind load.

The additional calculation device 63 uses an adder 204f to add the distortion due to gravity acting on the wind turbine blade 3, taking into account the azimuth angle, tilt angle, and cone angle.
Then, the additional calculation device 63 calculates the bending moment M _FLAP around the axis along the code C direction by multiplying the distortion moment correction coefficient (k1) by the multiplier 205f.

Similarly, the additional arithmetic unit 63 uses the subtractor 201e to subtract the bending strain correction coefficient ZERO adjustment (b2) from the strain εedge of the sensor 12D detected by the sensor 12D, and subtracts the strain corresponding to the thermal elongation of the wind turbine blade 3. The subtractor 202e performs subtraction.

The additional calculation device 63 subtracts the distortion due to centrifugal force using a subtractor 203e. Specifically, as described above, the additional calculation device 63 subtracts the tensile strain calculated by the multiplier 206 using the subtracter 203e.

The additional calculation device 63 uses an adder 204e to add the distortion due to gravity acting on the wind turbine blade 3, taking into account the azimuth angle, tilt angle, and cone angle.
Then, the additional calculation device 63 calculates the bending moment M _EDGE about the code C and the axis perpendicular to the central axis of the wind turbine blade 3 by multiplying the strain moment calibration coefficient (k2) by the multiplier 205e.

In this way, the additional calculation device 63 calculates the load acting on the wind turbine blade 3 based on the detection signal from the sensor 12A as the first sensor and the detection signal from the sensor 12D as the second sensor. Can be done.

Here, the wind turbine blade 3 is generally formed of fiber-reinforced plastic such as GFRP (glass fiber reinforced plastic) or CFRP (carbon fiber reinforced plastic). Therefore, when installing the sensor 12 to detect distortion of the wind turbine blade 3 at the manufacturing stage of the wind turbine blade 3, the work can be done on the ground, and it is relatively easy to embed the sensor 12 between the fiber layers. Therefore, installation of the sensor 12 is relatively easy.
However, when the sensor 12 is added to the wind turbine blade 3 of the existing wind power generation device 1, the wind turbine blade 3 remains attached to the wind turbine rotor 5, so the posture of the wind turbine blade 3 is different from that at the manufacturing stage of the wind turbine blade 3. Restrictions are imposed on the sensor 12, such as making it difficult to position the sensor 12 in an appropriate position. Therefore, installing the sensor 12 is a relatively difficult task, which involves arranging the sensor 12 on the wind turbine blade 3 and further covering and fixing the sensor 12 with fiber-reinforced plastic. Therefore, when additionally installing the sensors 12 on the wind turbine blades 3 of the existing wind power generation device 1, it is desirable that the number of additionally installed sensors 12 is small.

Therefore, as described above, the leeward bending moment per blade and the torque component per blade are calculated based on the detection signals from the two sensors attached to the blade roots 3a of each of the wind turbine blades 3. Then, the number of sensors 12 to be installed in the process of additionally installing the sensors 12 (additional installation process S10) is only two per blade. Therefore, the man-hours for additional installation work can be significantly reduced. This makes it possible to reduce the cost of adding the sensor 12 when providing a method for diagnosing the wind turbine blade 3 in the existing wind power generation device 1 by measuring the strain of the wind turbine blade 3 in real time.

When calculating the load acting on the wind turbine blade 3, it is preferable to calculate the load excluding the effects of centrifugal force acting on the wind turbine blade 3 and gravity acting on the wind turbine blade 3, as described above.
Thereby, the bending moment acting on the wind turbine blade 3 due to the wind load can be determined, with the effects of centrifugal force acting on the wind turbine blade 3 and gravity acting on the wind turbine blade 3 being eliminated.

(Computation contents based on detection signals from four optical fiber sensors 12A to 12D)
The bending moment M _FRAP around the axis along the code C direction acting on the wind turbine blade 3 and the bending moment M _EDGE around the axis perpendicular to the code C and the central axis (blade axis L0) of the wind turbine blade 3 are shown in FIG. 2A. A case will be described in which calculation is performed based on detection signals from four optical fiber sensors 12 (12A to 12D) as shown in FIG.

Bending moment M around the axis along the code C direction due to wind load Before and after _FRAP acts, the bending moment from two sensors (sensor 12D and sensor 12B) placed at two opposite positions along the code C direction is For each strain of the wind turbine blade 3 determined from the detection signal, the amount of change in strain is the same, but the increase/decrease in strain is opposite. Therefore, one-half of the difference between the two calculated amounts of change in strain becomes the amount of change in strain due to the bending moment M _FRAP . Further, when calculating the difference between the two amounts of change in strain obtained from the detection signals from the two sensors, the effects of centrifugal force and gravity acting on the wind turbine blades 3 are canceled out. Therefore, calculation of the bending moment M _FRAP becomes easy. The same can be said of the calculation of the bending moment M _EDGE around the axis perpendicular to the wind load code C and the central axis of the wind turbine blade 3 (blade axis L0).

The following description will be made based on the block diagram of FIG.
The additional calculation device 63 corrects the strain εflapA of the sensor 12A detected by the sensor 12A using the temperature correction unit 101f, thereby removing the influence of the thermal elongation of the wind turbine blade 3 from the strain εflapA. Similarly, the additional calculation device 63 removes the influence of thermal elongation of the wind turbine blade 3 from the strain εflapC by correcting the strain εflapC of the sensor 12C detected by the sensor 12C using the temperature correction unit 101f.
Note that the additional calculation device 63 omits the correction in the correction section 102f.

Then, the additional arithmetic unit 63 calculates the bending moment M _FRAP in the correction section 103f using the following equation (3) from the strain εflapA and the strain εflapC after removing the influence of thermal expansion in the temperature correction section 101f.
M _FRAP =k3×(εflapA−εflapC)/2−b3) ・・・(3)
Here, k3 is a strain moment calibration coefficient, which is a coefficient for converting strain into a bending moment.
b3 is the bending distortion correction coefficient ZERO adjustment, which is the distortion when the bending moment is zero.

Similarly, the additional calculation device 63 corrects the strain εflapD of the sensor 12D detected by the sensor 12D using the temperature correction unit 101e, thereby removing the influence of the thermal elongation of the wind turbine blade 3 from the strain εflapD. Similarly, the additional calculation device 63 removes the influence of thermal elongation of the wind turbine blade 3 from the strain εflapB by correcting the strain εflapB of the sensor 12B detected by the sensor 12B using the temperature correction unit 101e.
Note that the additional calculation device 63 omits the correction in the correction unit 102e.

Then, the additional arithmetic unit 63 calculates the bending moment M _EDGE in the correction unit 103e using the following equation (4) from the strain εflapD and the strain εflapB after removing the influence of thermal expansion in the temperature correction unit 101e.
M _EDGE =k4×(εflapD−εflapB)/2−b4) ・・・(4)
Here, k4 is a distortion moment calibration coefficient, and is a coefficient for converting distortion into a bending moment.
b4 is the bending distortion correction coefficient ZERO adjustment, which is the distortion when the bending moment is zero.

Although a detailed explanation of the processing based on FIG. 7 will be omitted, when performing calculations based on detection signals from the four optical fiber sensors 12 (12A to 12D) as shown in FIG. 2A, the additional calculation device 63 omits the subtraction in the subtracters 203f, 204f, and 203e.

(Calculation of bending moment acting on wind turbine rotor 5 and nacelle 4 by implementing coordinate transformation)
Hereinafter, calculation of the bending moment acting on the wind turbine rotor 5 and the nacelle 4 by executing coordinate transformation will be explained.
FIG. 8 is a block diagram for calculating the composite torque component MXR, that is, the rotor torque of the wind turbine rotor 5. As shown in FIG.
FIG. 9 is a block diagram for calculating the bending moment for the wind turbine rotor 5 and the bending moment for the nacelle 4.
FIG. 10A is a diagram for explaining the bending moment about the wind turbine blade 3.
FIG. 10B is a diagram for explaining the bending moment about the wind turbine rotor 5. FIG.
FIG. 10C is a diagram for explaining the bending moment regarding the nacelle 4.
The following calculations are performed by the existing calculation device 50 included in the existing wind power generation device 1, such as the hub PLC 51, or by the additional calculation device 63. For convenience of explanation, the following description assumes that each calculation is performed in the additional calculation device 63, but as described above, the calculations are performed in the existing calculation device 50 provided in the existing wind power generation device 1, such as the hub PLC 51. You can.

(Calculation of composite torque component MXR)
In addition, in the load calculation step S50 according to some embodiments, as described below, calculation processing of the composite torque component MXR, calculation processing of the bending moment about the wind turbine rotor 5, calculation processing of the bending moment about the nacelle 4, etc. Arithmetic processing may also be performed.
These calculation processes are performed by the existing calculation device 50 provided in the existing wind power generation device 1, such as the hub PLC 51, or by the additional calculation device 63. For convenience of explanation, the following explanation assumes that these calculation processes are performed in the additional calculation device 63. However, as described above, the existing calculation devices 50 included in the existing wind power generation device 1, such as the hub PLC 51, May be implemented.

As shown in FIG. 8, the additional calculation device 63 calculates the torque components per blade (first blade torque component, second blade torque component, A composite torque component MXR is calculated by adding the third blade torque component).

(Bending moment about wind turbine rotor 5)
As shown in FIG. 9, the additional calculation device 63 calculates the leeward bending moment per blade in each of the three wind turbine blades 3, calculated as described above, based on the mounting position of each wind turbine blade 3. By performing the correction and adding them together, the vertical resolved bending moment MYR and the horizontal resolved bending moment MZR for the wind turbine rotor 5 are calculated (see FIG. 10B).

(Bending moment about nacelle 4)
As shown in FIG. 9, the additional calculation device 63 corrects the vertical resolved bending moment MYR and the horizontal resolved bending moment MZR for the wind turbine rotor 5 by the azimuth angle, and adds them together. A composite vertical bending moment Md for the nacelle 4 is calculated (see FIG. 10C).
Further, the additional calculation device 63 corrects the vertical resolved bending moment MYR and the horizontal direction resolved bending moment MZR for the wind turbine rotor 5 using the azimuth angle, adds them together, and performs correction using the tilt angle. Then, a composite left-right bending moment Mq for the nacelle 4 is calculated (see FIG. 10C).

(About calibration of sensor 12)
The method for diagnosing a wind turbine blade according to some embodiments may include a calibration step S200 for the sensor 12, which will be described below.
FIG. 11 is a flowchart showing the processing procedure for the sensor 12 calibration step S200.
The calibration step S200 of the sensor 12 is a step of configuring the additionally installed sensor 12, and as shown in FIG. A first acquisition step S210 of acquiring the information, a second acquisition step S230 of acquiring the detection signal from the sensor at the azimuth angle where the blade axis line L0 comes closest in the vertical direction, and calibrating the sensor 12 based on the acquired detection signal. and a calibration step S250.
Note that the calibration step S200 of the sensor 12 according to this embodiment is performed by a worker.

The first acquisition step S210 is a step in which the wind turbine blade 3 is fixed at an azimuth angle where the blade axis L0 of the wind turbine blade 3 is horizontal, and a detection signal from the sensor 12 of the wind turbine blade 3 is acquired.
In the first acquisition step S210, the wind turbine blade 3 is fixed at an azimuth angle at which the blade axis L0 of the wind turbine blade 3 to which the sensor 12 to be calibrated is attached is horizontal, for example, at the 3 o'clock position when the wind turbine 1a is viewed from the front. do. Then, the pitch angle of the wind turbine blade 3 is changed, and the wavelength (detection signal) of the sensor 12 arranged along the code C direction is recorded at the pitch angle at which the wavelength of the sensor 12 becomes maximum or minimum. . Next, the pitch angle of the wind turbine blade 3 is changed, and the pitch angle at which the wavelength of the sensor 12 arranged along the direction orthogonal to the code C direction is maximum or minimum is determined as the wavelength of the sensor 12 (detection signal). Record.

Next, the wind turbine blade 3 is moved to the 9 o'clock position when the wind turbine 1a is viewed from the front, and then fixed. Then, as was done at the 3 o'clock position, the wavelength (detection signal) of the sensor 12 is recorded.

If the azimuth angle is plotted on the horizontal axis and the wavelength is plotted on the vertical axis, the wavelength curve will have a shape similar to a sine wave. Therefore, the zero point and amplitude are tentatively determined from the wavelength data recorded as described above.

The second acquisition step S230 is a step of acquiring a detection signal from the sensor 12 by fixing the wind turbine blade 3 at an azimuth angle where the blade axis L0 is closest to the vertical direction.
In the second acquisition step S230, the wind turbine blade 3 is fixed at the 12 o'clock position and the 6 o'clock position when the wind turbine 1a is viewed from the front, and the wavelength (detection signal) of the sensor 12 is recorded.
Note that the above-mentioned azimuth angle at which the blade axis L0 approaches the closest in the vertical direction is specifically the 12 o'clock position and the 6 o'clock position when the wind turbine 1a is viewed from the front. For example, if both the cone angle and the tilt angle are 0 degrees, the blade axis L0 points in the vertical direction at the 12 o'clock position and the 6 o'clock position when the wind turbine 1a is viewed from the front.

Further, the wind turbine blade 3 is fixed at a plurality of positions having arbitrary azimuth angles, and the wavelength (detection signal) of the sensor 12 is recorded as in the case of the 3 o'clock position. For example, it is preferable to record the wavelength (detection signal) of the sensor 12 while changing the azimuth angle in increments of 30 degrees.
Note that the order of recording the wavelengths (detection signals) of the sensor 12 at each azimuth angle as described above is not limited to the order described above, and may be performed in any order.
Furthermore, the wavelengths (detection signals) of the sensor 12 for each of the three wind turbine blades 3 may be recorded in parallel. For example, when the first wind turbine blade 31 is at the 3 o'clock position and the second wind turbine blade 32 is at the 11 o'clock position, first record the wavelength (detection signal) of the sensor 12 of the first wind turbine blade 31, and then Rather than recording the wavelength (detection signal) of the sensor 12 at another position of the first windmill blade 31, the second windmill blade 32 is moved to the 12 o'clock position and the wavelength (detection signal) of the sensor 12 of the second windmill blade 32 is recorded. signals) may be recorded.

The calibration step S250 is a step of calibrating the sensor 12 based on the wavelength data obtained as described above. Specifically, in the calibration step S250, the wavelength data obtained as described above is fitted to a sine wave by, for example, the least squares method. Then, assuming that the half amplitude of the sine wave obtained in this way matches the wavelength due to the bending moment due to the wind turbine blade 3's own weight, the above-mentioned distortion moment calibration coefficient and bending distortion calibration coefficient ZERO adjustment are determined. .

Note that if the additional sensor is the sensor 42, "wavelength" in the above description will be replaced with "distortion".

The wind turbine blade diagnosis method according to some embodiments includes the above-described sensor 12 calibration step S200, so that even if a phenomenon in which the output value of the sensor 12 deviates over time, that is, a so-called drift occurs, The influence of drift of the sensor 12 can be reduced.

According to the wind turbine blade diagnosis method according to some embodiments, by fixing the wind turbine blade 3 at an azimuth angle where the blade axis L0 of the wind turbine blade 3 is horizontal and acquiring a detection signal from the sensor 12, A detection signal based on the bending moment acting on the wind turbine blade 3 due to its own weight can be obtained.
By fixing the wind turbine blade 3 at the azimuth angle where the blade axis L0 is closest to the vertical direction and acquiring the detection signal from the sensor 12, the influence of the bending moment acting on the wind turbine blade 3 due to its own weight can be reduced. It is possible to obtain a detection signal that eliminates as much as possible.
In addition, in each process, the wind turbine blade 3 is fixed so that the wind turbine rotor 5 does not rotate, and the detection signal from the sensor 12 is acquired, so the detection signal is obtained from which the influence of acceleration and deceleration of the rotation of the wind turbine blade 3 has been eliminated. Can be obtained.
Thereby, the calibration accuracy of the sensor 12 can be improved.

(About detecting the presence or absence of abnormality in wind turbine blades)
Hereinafter, the contents of specific calculation processing in the abnormality detection step S90 according to some embodiments will be explained.
The calculation processing described below is performed by the existing calculation device 50 included in the existing wind power generation device 1, such as the hub PLC 51, or the additional calculation device 63. For convenience of explanation, the following explanation assumes that the additional calculation device 63 performs the calculation processing described below. It may also be implemented by the device 50.

Based on the detection signal from the sensor 12 additionally installed as described above, it is possible to detect the presence or absence of an abnormality in the wind turbine blade 3, such as damage to the wind turbine blade 3.
For example, if delamination occurs between layers in the fiber-reinforced plastic that constitutes the wind turbine blade 3, the tendency of the strain fluctuation that appears during one rotation of the wind turbine blade 3 may change, the range of strain fluctuation, and the average of the fluctuating strain may change. Changes appear in values, etc.
By detecting such changes, it is possible to detect whether or not there is an abnormality in the wind turbine blades 3.

FIG. 12 is an example of a schematic cross-sectional view of a wind turbine blade 3 according to some embodiments in a cross section perpendicular to the blade axis L0. FIG. 12 shows a cross section at a position away from the base end of the wind turbine blade 3 toward the outside in the radial direction of the wind turbine rotor 5, and the cross-sectional shape of the wind turbine blade 3 is a slightly flattened cylindrical shape.
In the wind turbine blade 3 according to some embodiments, as shown in FIG. 12, a reinforcing member 301 for reinforcing the wind turbine blade 3 is provided inside the wind turbine blade 3 having a cylindrical shape. In the example shown in FIG. 12, the reinforcing member 301 is attached to the inner circumferential surface 3b of the wind turbine blade 3 along the blade axis line so as to divide the inside of the wind turbine blade 3 into a plurality of regions when viewed along the blade axis line L0. It extends from a region on one side of L0 (for example, the lower region in the figure) to a region on the other side (for example, the upper region in the figure).
In the example shown in FIG. 12, both ends 301a of the reinforcing member 301 are connected to the inner circumferential surface 3b of the wind turbine blade 3 by bonding or the like.

(Damage detection by comparison between wind turbine blades 3)
For example, if both ends 301a of the reinforcing member 301 in one wind turbine blade 3 out of a plurality of wind turbine blades 3 are separated from the inner circumferential surface 3b of the wind turbine blade 3, other wind turbine blades 3 where such separation has not occurred There will be a difference in the magnitude of strain when a load is applied.
Therefore, damage to the wind turbine blades 3 can be detected by comparing the strains measured by the optical fiber sensor 12 or the sensor 42 between the wind turbine blades 3 as described above.

That is, the additional calculation device 63 detects, for example, the wind turbine blade based on the detection signal from the sensor (for example, the optical fiber sensor 12 or the sensor 42) for detecting the distortion of the wind turbine blade 3, which is acquired in the detection signal acquisition step. Compare the distortion among the three. Then, if the difference in the compared distortions is equal to or greater than a predetermined threshold value, the additional calculation device 63 may determine that an abnormality has occurred in the wind turbine blade 3.
Here, for example, the sensors 12 and 42 additionally installed in the additional installation step S10 include a first wind turbine blade sensor (sensor 12, 42 ) and a second wind turbine blade sensor (sensors 12, 42) for detecting distortion of a second wind turbine blade 32 different from the first wind turbine blade 31 among the plurality of wind turbine blades.
In addition, in the detection signal acquisition step S30, the light source/signal processing unit 10 or the strain gauge signal conditioner 11 receives the first wind turbine blade detection signal from the first wind turbine blade sensor (sensors 12, 42) and the second wind turbine blade. It is preferable to obtain the second wind turbine blade detection signal from the sensor (sensor 12, 42).
In the abnormality detection step S90, the additional calculation device 63 detects the first wind turbine blade detection signal and the second wind turbine blade detection signal based on the first wind turbine blade detection signal and the second wind turbine blade detection signal. It is preferable to detect damage to the first wind turbine blade 31, that is, the presence or absence of an abnormality, based on a converted signal output from the light source/signal processing unit 10 or the strain gauge signal conditioner 11.

In the abnormality detection step S90, the additional calculation device 63 calculates the distortion of the first wind turbine blade 31 and the second wind turbine blade 32 as described above based on the conversion signal, and then detects the abnormality of the first wind turbine blade 31. The presence or absence may be detected, or the presence or absence of an abnormality in the first wind turbine blade 31 may be detected based on the conversion signal without calculating the distortion of the first wind turbine blade 31 and the second wind turbine blade 32.

Thereby, the presence or absence of an abnormality in the first wind turbine blade 31 can be easily detected by referring to the distortion of the second wind turbine blade 32, which is different from the first wind turbine blade 31, which is the target for detecting the presence or absence of an abnormality.

(Damage detection without comparison between wind turbine blades 3)
As described above, before and after a bending moment along the code C direction acts on the wind turbine blade 3 due to wind load, the distortion of the wind turbine blade determined from the detection signals from the leading edge side sensor and the trailing edge side sensor, respectively. The amount of change in distortion is the same, but the increase/decrease in distortion is opposite.
Therefore, if the above relationship no longer holds true between the detection signal from the leading edge side sensor and the detection signal from the trailing edge side sensor, it is likely that some kind of problem has occurred in the wind turbine blade 3, unless there is a problem with the sensor 12. It will be done.
Similarly, if the above relationship no longer holds true between the detection signal from the ventral sensor and the detection signal from the dorsal sensor, this indicates that some kind of malfunction has occurred in the wind turbine blade 3, unless there is a malfunction in the sensor 12. Conceivable.

Therefore, for example, as shown in FIGS. 1 and 2A, a leading edge side sensor (optical fiber sensor 12D), a trailing edge side sensor (optical fiber sensor 12B), a ventral side sensor (optical fiber sensor 12A), and a dorsal side sensor (optical fiber sensor 12D), If the sensor 12C) is included, the presence or absence of an abnormality in the wind turbine blade 3 may be detected in the following manner.
In the detection signal acquisition step S30, the light source/signal processing unit 10 includes a leading edge side sensor (optical fiber sensor 12D), a trailing edge side sensor (optical fiber sensor 12B), a ventral side sensor (optical fiber sensor 12A), and a dorsal side sensor (optical fiber sensor 12D). It is preferable to acquire the second wind turbine blade detection signal from the fiber sensor 12C). Note that four sensors 42 may be used instead of these four sensors 12.

In the abnormality detection step S90, the additional calculation device 63 may detect the presence or absence of an abnormality in the wind turbine blade 3 based on the comparison result of at least one of the following (a) or (b).
(a) Comparison result between the detection signal from the leading edge sensor (optical fiber sensor 12D) and the detection signal from the trailing edge sensor (optical fiber sensor 12B) (more specifically, the comparison result with the conversion signal derived from these detection signals) Comparison result).
(b) Comparison result between the detection signal from the ventral sensor (optical fiber sensor 12A) and the detection signal from the dorsal sensor (optical fiber sensor 12C) (more specifically, the comparison result with the conversion signal derived from these detection signals) Comparison result).

In the abnormality detection step S90, the additional calculation device 63 may detect the presence or absence of an abnormality in the wind turbine blade 3 based on the conversion signal without calculating the distortion of the wind turbine blade 3; After calculating the distortion of the wind turbine blade 3 as described above, the presence or absence of an abnormality in the wind turbine blade 3 may be detected.

As a result, the detection signal from the leading edge sensor (optical fiber sensor 12D) and the detection signal from the trailing edge sensor (optical fiber sensor 12B), or the detection signal from the ventral sensor (optical fiber sensor 12A) and the dorsal sensor The presence or absence of an abnormality in the wind turbine blade 3 can be detected based on at least one of the detection signals from the (optical fiber sensor 12C).

In another embodiment, the additional calculation device 63 receives a detection signal from a sensor (for example, the optical fiber sensor 12 or the sensor 42) for detecting the distortion of the wind turbine blade 3, which is acquired in the detection signal acquisition step. Based on this, changes in strain over time are stored in a storage device (not shown) over a relatively long period of time, such as several months or years. If the amount of change in the strain over time as described above is greater than or equal to a predetermined threshold value, the additional calculation device 63 may determine that an abnormality has occurred in the wind turbine blade 3.

When it is determined that an abnormality has occurred in the wind turbine blade 3 as described above, for example, the additional calculation device 63 informs a device external to the wind power generation device 1 via the communication device 67 that an abnormality has occurred in the wind turbine blade 3. A signal indicating that the vehicle is present may be transmitted wirelessly.

(Estimation of azimuth angle)
In some of the embodiments described above, the information on the azimuth angle was acquired from the existing wind power generation device 1. However, a sensor for detecting the azimuth angle may be additionally installed, and information on the azimuth angle may be acquired from the additionally installed sensor.

For example, the additional installation step S10 described above may include a step of additionally installing an azimuth angle estimation sensor for estimating the azimuth angle of the wind turbine blade 3. In this step, the azimuth angle estimation sensor 44 (see FIG. 1) is attached to any member that rotates together with the wind turbine rotor 5, such as the hub 2.
The azimuth angle estimation sensor 44 preferably includes one acceleration sensor that can detect accelerations in three axes: the X-axis, the Y-axis, and the Z-axis. The acceleration sensor may detect the direction in which gravity acts, and the additional calculation device 63 may estimate the azimuth angle from the detection result of the acceleration sensor.
Note that if the azimuth angle estimation sensor 44 also includes one angular velocity sensor capable of detecting angular velocities around the three axes of the X-axis, Y-axis, and Z-axis, the acceleration sensor If the direction of gravity cannot be detected from the detection result, the azimuth angle may be estimated as follows. In other words, the direction of gravity before it becomes impossible to detect the direction of gravity due to vibrations, etc., the angular velocity of the wind turbine rotor 5 detected by the angular velocity sensor, and the elapsed time after the direction of gravity can no longer be detected due to vibrations, etc. The additional calculation device 63 may estimate the azimuth angle based on .
Thereby, even if azimuth angle information cannot be obtained from the existing wind power generator 1, azimuth angle information can be obtained by the azimuth angle estimation sensor 44.

(Pitch angle estimation)
In some of the embodiments described above, the pitch angle information was acquired from the existing wind power generation device 1. However, a sensor for detecting the pitch angle may be additionally installed, and pitch angle information may be acquired from the additional sensor.

For example, the additional installation step S10 described above may include a step of additionally installing a pitch angle estimation sensor for estimating the pitch angle of the wind turbine blade. In this step, the pitch angle estimation sensor 46 (see FIG. 1) may be added to each of the plurality of wind turbine blades 3, for example. Note that if the difference in pitch angle between the plurality of wind turbine blades 3 is not a problem, the pitch angle estimation sensor 46 may be additionally installed on any one of the plurality of wind turbine blades 3 in this step. .
The pitch angle estimation sensor 46 preferably includes one acceleration sensor that can detect accelerations in three axes: the X-axis, the Y-axis, and the Z-axis.
When the wind turbine blade 3 is at the 3 o'clock and 6 o'clock positions, the direction in which gravity acts is orthogonal to the blade axis L0 of the wind turbine blade 3. Therefore, the acceleration sensor detects the direction in which gravity acts at the 3 o'clock and 6 o'clock positions, and the detection result of the acceleration sensor and the pitch angle and gravity that have been measured in advance at the 3 o'clock and 6 o'clock positions are used. The pitch angle may be estimated by the additional calculation device 63 from the relationship with the direction of action.
Thereby, even if pitch angle information cannot be obtained from the existing wind power generator 1, pitch angle information can be obtained by the pitch angle estimation sensor 46.

The present disclosure is not limited to the embodiments described above, and also includes forms in which modifications are added to the embodiments described above, and forms in which these forms are appropriately combined.

The contents described in each of the above embodiments can be understood as follows, for example.
(1) A method for diagnosing a wind turbine blade according to at least one embodiment of the present disclosure is a method for diagnosing a wind turbine blade 3 of an existing wind power generation device 1, and includes sensors 12 and 42 for detecting distortion of the wind turbine blade 3. (additional installation step S10), a step of acquiring detection signals from the sensors 12 and 42 (detection signal acquisition step S30), and a step of calculating the load acting on the wind turbine blade 3 based on the detection signals. (Load calculation step S50). In the step of additionally installing the sensors 12 and 42 (additional installation step S10), sensors 12 and 42 are placed oppositely along the code C direction of the wind turbine blade 3 among the circumferential positions in the blade root portion 3a centered on the blade axis L0 of the wind turbine blade 3. A first sensor (sensors 12, 42) is placed at one of two positions, and a second sensor (sensors 12, 42) is placed at one of two positions facing each other along the direction orthogonal to the code C direction. Set up. In the step of calculating the load (load calculation step S50), the load is calculated based on the detection signal from the first sensor (sensors 12, 42) and the detection signal from the second sensor (sensor 12, 42).

For example, sensors 12 for detecting distortion of the wind turbine blade 3 are additionally installed at two positions facing each other along the code C direction and at two positions facing each other along a direction orthogonal to the code C direction. Think about the case.
Bending moment M around the axis along the code C direction due to wind load Before and after _FRAP acts, the bending moment from two sensors (sensor 12D and sensor 12B) placed at two opposite positions along the code C direction is For each strain of the wind turbine blade 3 determined from the detection signal, the amount of change in strain is the same, but the increase/decrease in strain is opposite. Therefore, one-half of the difference between the two calculated amounts of change in strain becomes the amount of change in strain due to the bending moment M _FRAP . Further, when calculating the difference between the two amounts of change in strain obtained from the detection signals from the two sensors, the effects of centrifugal force and gravity acting on the wind turbine blades 3 are canceled out. Therefore, calculation of the bending moment M _FRAP becomes easy. The same can be said of the calculation of the bending moment M _EDGE around the axis perpendicular to the wind load code C and the central axis of the wind turbine blade 3 (blade axis L0).

Here, the wind turbine blade 3 is generally formed of fiber-reinforced plastic such as GFRP (glass fiber reinforced plastic) or CFRP (carbon fiber reinforced plastic). Therefore, when installing the sensor 12 to detect distortion of the wind turbine blade 3 during the manufacturing stage of the wind turbine blade 3, the work can be done on the ground, and it is relatively easy to embed the sensor 12 between the fiber layers. Therefore, installation of the sensor 12 is relatively easy.
However, when the sensor 12 is added to the wind turbine blade 3 of the existing wind power generation device 1, the wind turbine blade 3 remains attached to the wind turbine rotor 5, so the posture of the wind turbine blade 3 is different from that at the manufacturing stage of the wind turbine blade 3. Restrictions such as difficulty in setting the sensor 12 in a suitable posture are imposed on the sensor 12. Therefore, installing the sensor 12 is a relatively difficult task, which involves arranging the sensor 12 on the wind turbine blade 3 and further covering and fixing the sensor 12 with fiber-reinforced plastic. Therefore, when additionally installing sensors 12 on the wind turbine blades 3 of the existing wind power generation device 1, it is desirable that the number of additionally installed sensors 12 is small.

According to the method (1) above, the number of sensors 12 and 42 to be installed in the step of adding the sensors 12 and 42 (additional installation step S10) is only two, so the man-hours for the additional installation work are significantly reduced. can. This makes it possible to reduce the cost of additionally installing the sensors 12 and 42 when providing a method for diagnosing the wind turbine blades 3 in the existing wind power generation device 1 by measuring the strain of the wind turbine blades 3 in real time.

(2) In some embodiments, in the method of (1) above, in the step of calculating the load (load calculation step S50), the centrifugal force acting on the wind turbine blade 3 and the gravitational force acting on the wind turbine blade 3 are It is better to calculate the load excluding the influence.

As described above, in the method (1) above, in the step of additionally installing the sensors 12 and 42 (additional installation step S10), one of the two positions facing each other along the code C direction, and Two sensors 12 and 42 are additionally provided, which are placed at one of two opposing positions along the direction perpendicular to the direction. Therefore, in order to determine the bending moment that acts on the wind turbine blade 3 due to the wind load, it is necessary to eliminate the effects of centrifugal force that acts on the wind turbine blade 3 and gravity that acts on the wind turbine blade.
According to the method (2) above, since the load is calculated excluding the effects of centrifugal force acting on the wind turbine blade 3 and gravity acting on the wind turbine blade 3, the wind load acts on the wind turbine blade 3. Bending moment can be determined.

(3) In some embodiments, in the method (2) above, in the step of calculating the load (load calculation step S50), the wind turbine blade is The centrifugal force acting on the wind turbine blade 3 obtained from the detection signals from the sensors 12 and 42 obtained at the azimuth angle where the blade axis L0 is closest to the vertical direction, and the rotational speed of the wind turbine rotor 5 equipped with the wind turbine blade 3. A load that excludes the influence of centrifugal force acting on the wind turbine blade 3 may be calculated based on a relationship obtained in advance with respect to the square of the wind turbine blade.

According to the method (3) above, the influence of centrifugal force acting on the wind turbine blade 3 can be eliminated when determining the bending moment acting on the wind turbine blade 3 due to wind load.

(4) A method for diagnosing a wind turbine blade according to at least one embodiment of the present disclosure is a method for diagnosing a wind turbine blade 3 of an existing wind power generation device 1, and includes sensors 12 and 42 for detecting distortion of the wind turbine blade 3. (additional installation step S10), a step of acquiring detection signals from the sensors 12 and 42 (detection signal acquisition step S30), and a step of calculating the load acting on the wind turbine blade 3 based on the detection signals. (load calculation step S50); and a step of detecting whether or not there is an abnormality in the wind turbine blade 3 based on the detection signal (abnormality detection step S90).

According to the method (4) above, the presence or absence of an abnormality in the wind turbine blade 3 can be detected based on the distortion of the wind turbine blade 3.

(5) In some embodiments, in the method of (4) above, the sensors 12, 42 added in the step of adding the sensors 12, 42 (addition step S10) detect the presence or absence of an abnormality. Therefore, it is preferable to include an abnormality detection sensor that detects distortion of the wind turbine blades 3. In the step of detecting the presence or absence of an abnormality (abnormality detection step S90), the presence or absence of an abnormality may be detected based on a detection signal from an abnormality detection sensor.

According to the method (5) above, if the detection signal from the abnormality detection sensor is not required when calculating the load acting on the wind turbine blade 3, the processing of the detection signal can be simplified, and the sensor 12, 42 In the arithmetic device for processing the detection signal from the sensor, the load can be suppressed.

(6) In some embodiments, in the method (4) or (5) above, the existing wind power generation device 1 has a wind turbine rotor 5 including a plurality of wind turbine blades 3. The sensors 12 and 42 additionally installed in the step of additionally installing the sensor 12 (addition step S10) include a first sensor (first Wind turbine blade sensors (sensors 12, 42)) and a second sensor (second wind turbine blade sensor (sensor 12, 42)). In the step of acquiring a detection signal (detection signal acquisition step S30), the first detection signal (first wind turbine blade detection signal) from the first sensor (first wind turbine blade sensor (sensor 12, 42)) and the second sensor It is preferable to obtain a second detection signal (second wind turbine blade detection signal) from (second wind turbine blade sensor (sensor 12, 42)). In the step of detecting the presence or absence of an abnormality (abnormality detection step S90), the detection of the first wind turbine blade 31 is performed based on the first detection signal (first wind turbine blade detection signal) and the second detection signal (second wind turbine blade detection signal). It is good to detect the presence or absence of an abnormality.

According to the method (6) above, the presence or absence of an abnormality in the first wind turbine blade 31 is detected by referring to the distortion of the second wind turbine blade 32, which is different from the first wind turbine blade 31, which is the target for detecting the presence or absence of an abnormality. becomes easier.

(7) In some embodiments, in the method (4) or (5) above, the sensors 12, 42 added in the step of adding the sensors 12, 42 (addition step S10) include wind turbine blades. Among the circumferential positions in the blade root portion 3a centered on the blade axis L0 of the wind turbine blade 3, the leading edge sides are respectively arranged at two positions, the leading edge side 23 and the trailing edge side 24, which face each other along the code C direction of the wind turbine blade 3. A sensor (optical fiber sensor 12D), a trailing edge sensor (optical fiber sensor 12B), and ventral sensors arranged at two positions, ventral side 21 and dorsal side 22, facing each other along the direction orthogonal to the code C direction. A side sensor (optical fiber sensor 12A) and a back sensor (optical fiber sensor 12C) may be included. In the process of acquiring detection signals, a detection signal from the leading edge side sensor (optical fiber sensor 12D), a detection signal from the trailing edge side sensor (optical fiber sensor 12B), and a detection signal from the ventral side sensor (optical fiber sensor 12A) are acquired. It is preferable to acquire the signal and the detection signal from the dorsal sensor (optical fiber sensor 12C). In the step of detecting the presence or absence of an abnormality (abnormality detection step S90), the comparison result between the detection signal from the leading edge side sensor (optical fiber sensor 12D) and the detection signal from the trailing edge side sensor (optical fiber sensor 12B), or The presence or absence of an abnormality in the wind turbine blade 3 is detected based on the comparison result of at least one of the detection signal from the side sensor (optical fiber sensor 12A) and the detection signal from the back sensor (optical fiber sensor 12C). It's good to do that.

As described above, before and after a bending moment along the code C direction acts on the wind turbine blade 3 due to wind load, the distortion of the wind turbine blade determined from the detection signals from the leading edge side sensor and the trailing edge side sensor, respectively. The amount of change in distortion is the same, but the increase/decrease in distortion is opposite.
Therefore, if the above relationship no longer holds true between the detection signal from the leading edge side sensor and the detection signal from the trailing edge side sensor, it is likely that some kind of problem has occurred in the wind turbine blade 3, unless there is a problem with the sensor 12. It will be done.
Similarly, if the above relationship no longer holds true between the detection signal from the ventral sensor and the detection signal from the dorsal sensor, this indicates that some kind of malfunction has occurred in the wind turbine blade 3, unless there is a malfunction in the sensor 12. Conceivable.
According to the method (7) above, the detection signal from the leading edge side sensor (optical fiber sensor 12D) and the detection signal from the trailing edge side sensor (optical fiber sensor 12B), or the detection signal from the ventral side sensor (optical fiber sensor 12A) The presence or absence of an abnormality in the wind turbine blade 3 can be detected based on at least one of the detection signal from the back side sensor (optical fiber sensor 12C).

(8) In some embodiments, in any of the methods (1) to (7) above, the load information calculated in the step of calculating the load (load calculation step S50) is transmitted to the outside of the wind power generator 1. (external transmission step S70).

If the information on the load calculated in the step of calculating the load (load calculation step S50) is stored in a storage device etc. arranged inside the wind turbine generator 1, for example, if the information on the load is stored in the storage device etc. If an accident such as a collapse occurs and the storage device is destroyed, the information cannot be retrieved from the storage device, and the information cannot be used to investigate the cause of the accident.
According to method (8) above, if the above information is stored in a storage device etc. external to the wind power generator 1, even if an accident like the one described above occurs, the above information can be used to investigate the cause of the accident. can be used. That is, according to the method (8) above, the possibility of the information being lost can be reduced.

(9) In some embodiments, in any of the methods (1) to (8) above, in the step of calculating the load (load calculation step S50), based on the detection signal, the existing wind power generation device 1 The load may be calculated using an existing calculation device 50 provided in the vehicle.

According to the method (9) above, the number of additionally installed calculation units 63 can be suppressed, thereby providing a method of diagnosing the wind turbine blade 3 by measuring the strain of the wind turbine blade 3 in real time in the existing wind power generation device 1. In doing so, the cost of adding an arithmetic device can be reduced.

(10) In some embodiments, in any of the methods (1) to (8) above, a method for calculating a load acting on the wind turbine blade 3 on the wind turbine generator 1 based on a detection signal is provided. The process may include a step of adding an arithmetic device (addition step S10). In the step of calculating the load (load calculation step S50), the load may be calculated based on the detection signal using the calculation device 63 added in the step of adding a calculation device (additional step S10).

In general, if the existing calculation device 50 included in the existing wind power generation device 1 is not intended to calculate the load acting on the wind turbine blade 3 based on the detection signals from the sensors 12 and 42, the The capacity may not be sufficient to calculate the load.
According to the method (10) above, it is possible to additionally install the computing device 63 having sufficient computing power to compute the load.

(11) In some embodiments, in the method (10) above, power is supplied to the calculation device 63 added to the wind power generation device 1 in the step of adding the calculation device (addition step S10). It is preferable to include a step of additionally installing a possible power storage device 65 (additional installation step S10).

According to the method (11) above, even if the power supply from the wind power generator 1 side is cut off due to some kind of malfunction, the arithmetic device can be operated with the power from the power storage device 65.

(12) In some embodiments, any of the methods (1) to (11) above may include a step S200 of calibrating the sensor.

According to the method (12) above, even if a phenomenon in which the output values of the sensors 12 and 42 shift over time, so-called drift, occurs, the influence of the drift of the sensors 12 and 42 can be reduced.

(13) In some embodiments, in the method of (12) above, in step S200 of calibrating the sensor, the wind turbine blade 3 is fixed at least at an azimuth angle such that the blade axis L0 of the wind turbine blade 3 is horizontal. , a step S210 of acquiring detection signals from the sensors 12 and 42, and a step S230 of acquiring detection signals from the sensors 12 and 42 by fixing the wind turbine blade 3 at an azimuth angle where the blade axis line L0 is closest to the vertical direction. It is preferable to calibrate the sensors 12 and 42 based on the detection signals from the sensors 12 and 42 obtained by performing .

According to the method (13) above, the wind turbine blade 3 is fixed at an azimuth angle such that the blade axis L0 of the wind turbine blade 3 is horizontal, and the detection signals from the sensors 12 and 42 are acquired. A detection signal based on the bending moment acting on the wind turbine blade 3 due to its own weight can be obtained.
By fixing the wind turbine blade 3 at an azimuth angle where the blade axis L0 is closest to the vertical direction and acquiring detection signals from the sensors 12 and 42, the bending moment acting on the wind turbine blade 3 due to its own weight can be reduced. It is possible to obtain a detection signal that eliminates the influence of
In addition, in each process, the wind turbine blade 3 is fixed so that the wind turbine rotor 5 does not rotate, and the detection signals from the sensors 12 and 42 are acquired, so the influence of acceleration and deceleration of the rotation of the wind turbine blade 3 is eliminated. I can get the signal.
Thereby, the calibration accuracy of the sensors 12 and 42 can be improved.

(14) In some embodiments, in any of the methods (1) to (13) above, the step of additionally installing an azimuth angle estimation sensor for estimating the azimuth angle of the wind turbine blade 3; a step of additionally installing a pitch angle estimation sensor for estimating the pitch angle of the pitch angle, a step of estimating the azimuth angle based on a detection signal from the azimuth angle estimation sensor, and a detection signal from the pitch angle estimation sensor estimating the pitch angle based on the pitch angle.

According to the method (14) above, even if information on the azimuth angle and pitch angle cannot be obtained from the existing wind power generator 1, the azimuth angle and pitch angle can be determined by the azimuth angle estimation sensor and the pitch angle estimation sensor. Information can be obtained.

(15) In some embodiments, in any of the methods (1) to (14) above, in the step of acquiring a detection signal (detection signal acquisition step S30), the detection is performed based on the rotational speed of the wind turbine blade 3. The sampling frequency for acquiring the signal may be changed.

If the sampling frequency for acquiring detection signals is constant regardless of the rotational speed of the windmill blades 3, the larger the rotational speed of the windmill blades 3, the greater the azimuth angle from acquisition of a detection signal to acquisition of the next detection signal. The amount of change increases. Therefore, as the rotational speed of the wind turbine blade 3 increases, the number of detection signals that can be obtained during one rotation of the wind turbine blade 3 decreases, and fluctuations in the bending moment acting on the wind turbine blade 3 due to wind load are detected. Accuracy will decrease.
According to the method (15) above, if the sampling frequency for acquiring the detection signal is increased as the rotational speed of the windmill blade 3 increases, the accuracy of detecting the above fluctuation even if the rotational speed of the windmill blade 3 increases. can suppress the decline in

1 Wind power generator 1a Wind turbine 2 Hub 3 Wind turbine blade 4 Nacelle 5 Wind turbine rotor 10 Light source/signal processing unit 11 Strain gauge signal conditioner 12 Optical fiber sensor (sensor)
42 Sensor 50 Existing calculation device 51 Hub PLC
53 Nacelle PLC
55 Windmill PLC
57 Hub power supply 63 Computing device (additional computing device)
65 Power storage device 67 Communication device

Claims (12)

A method for diagnosing wind turbine blades of an existing wind power generation device, the method comprising:
a step of additionally installing a sensor for detecting distortion of the wind turbine blade;
acquiring a detection signal from the sensor;
calculating a load acting on the wind turbine blade based on the detection signal;
Detecting the presence or absence of an abnormality in the wind turbine blade based on the detection signal;
Equipped with
The sensor added in the step of adding the sensor includes an abnormality detection sensor that detects distortion of the wind turbine blade in order to detect the presence or absence of the abnormality,
In the step of detecting the presence or absence of the abnormality, detecting the presence or absence of the abnormality based on the detection signal from the abnormality detection sensor ,
In the step of calculating the load, the detection signal from the sensor is acquired at an azimuth angle where the blade axis of the wind turbine blade approaches the vertical direction most during a period before the existing wind power generation device starts generating electricity. The influence of the centrifugal force acting on the wind turbine blade was eliminated based on the relationship obtained in advance between the centrifugal force acting on the wind turbine blade obtained from Calculate the load
How to diagnose wind turbine blades. In the step of detecting the presence or absence of the abnormality, it is determined that the abnormality exists if the amount of change in the strain of the wind turbine blade over time based on the detection signal from the abnormality detection sensor is greater than or equal to a predetermined threshold. The method for diagnosing a wind turbine blade according to claim 1. The existing wind power generation device has a wind turbine rotor including a plurality of the wind turbine blades,
The sensors additionally installed in the step of additionally installing the sensors include a first sensor for detecting the distortion of the first wind turbine blade among the plurality of wind turbine blades, and a first sensor for detecting the strain of the first wind turbine blade among the plurality of wind turbine blades. a second sensor for detecting distortion of a second wind turbine blade different from the first wind turbine blade;
In the step of acquiring the detection signal, acquiring a first detection signal from the first sensor and a second detection signal from the second sensor,
The method for diagnosing a wind turbine blade according to claim 1 or 2, wherein in the step of detecting the presence or absence of an abnormality, the presence or absence of an abnormality in the first wind turbine blade is detected based on the first detection signal and the second detection signal. . The sensor additionally installed in the step of additionally installing the sensor includes the leading edge side and the rear side, which are opposite to each other along the chord direction of the wind turbine blade, among the circumferential positions at the blade root centering on the blade axis of the wind turbine blade. A leading edge side sensor and a trailing edge side sensor respectively arranged at two positions on the edge side, and a ventral side arranged at two positions on the ventral side and dorsal side facing each other along a direction orthogonal to the cord direction. a sensor and a dorsal sensor;
In the step of acquiring the detection signals, a detection signal from the leading edge side sensor, a detection signal from the trailing edge side sensor, a detection signal from the ventral side sensor, and a detection signal from the dorsal side sensor are acquired. ,
In the step of detecting the presence or absence of an abnormality, the comparison result between the detection signal from the leading edge side sensor and the detection signal from the trailing edge side sensor, or the detection signal from the ventral side sensor and the detection signal from the dorsal side sensor. The method for diagnosing a wind turbine blade according to claim 1 or 2, wherein the presence or absence of an abnormality in the wind turbine blade is detected based on at least one of the comparison results. a step of transmitting information on the load calculated in the step of calculating the load to the outside of the wind power generator;
The method for diagnosing a wind turbine blade according to any one of claims 1 to 4 , comprising: The method for diagnosing a wind turbine blade according to any one of claims 1 to 5 , wherein in the step of calculating the load, the load is calculated using an existing calculation device included in an existing wind power generation device based on the detection signal. . adding a calculation device to the wind power generation device for calculating a load acting on the wind turbine blade based on the detection signal;
Equipped with
The wind turbine according to any one of claims 1 to 6 , wherein in the step of calculating the load, the load is calculated based on the detection signal by the calculation device added in the step of adding the calculation device. How to diagnose wings. a step of adding to the wind power generation device a power storage device capable of supplying power to the arithmetic device added in the step of adding the arithmetic device;
The method for diagnosing a wind turbine blade according to claim 7 , comprising: calibrating the sensor;
The method for diagnosing a wind turbine blade according to any one of claims 1 to 8 , comprising: In the step of calibrating the sensor,
at least,
fixing the wind turbine blade at an azimuth angle such that the blade axis of the wind turbine blade is horizontal, and acquiring a detection signal from the sensor;
fixing the wind turbine blade at an azimuth angle where the blade axis is closest to the vertical direction, and acquiring a detection signal from the sensor;
The method for diagnosing a wind turbine blade according to claim 9 , wherein the sensor is calibrated based on a detection signal from the sensor obtained by performing. a step of additionally installing an azimuth angle estimation sensor for estimating the azimuth angle of the wind turbine blade;
a step of additionally installing a pitch angle estimation sensor for estimating the pitch angle of the wind turbine blade;
estimating the azimuth angle based on a detection signal from the azimuth angle estimation sensor;
estimating the pitch angle based on a detection signal from the pitch angle estimation sensor;
The method for diagnosing a wind turbine blade according to any one of claims 1 to 10 . The method for diagnosing a wind turbine blade according to any one of claims 1 to 11 , wherein in the step of acquiring the detection signal, a sampling frequency for acquiring the detection signal is changed based on the rotational speed of the wind turbine blade.

Images