Real-time monitoring and correction method of initial reference wavelength of blade root load sensor of wind turbine

Abstract

The present invention discloses a method for real-time monitoring and correction of an initial reference wavelength of a blade root load sensor of a power generation wind turbine, comprising the following steps: S1, mounting a grating optical fiber sensor on a blade, and calibrating the initial reference wavelength and sensitivity coefficient of the grating optical fiber sensor; S2, collecting the real-time wavelength of the grating optical fiber sensor; S3, calculating the real-time load of the blade based on the real-time wavelength, the initial reference wavelength and the sensitivity coefficient; S4, judging whether it is necessary to correct the initial reference wavelength of the grating optical fiber sensor based on load numerical logic; compared with the prior art, the present invention can reduce the problem of initial reference wavelength offset by correcting the initial reference wavelength of the grating optical fiber sensor, realize real-time correction, improve the load measurement accuracy, and reduce the time loss of repeated calibration.

Description

Technical Field

[0001] The invention relates to the field of wind turbine blade load correction, and in particular to a method for real-time monitoring and correction of an initial reference wavelength of a blade root load sensor of a power generation wind turbine. Background Art

[0002] Wind energy is a very abundant renewable energy source in nature, and its commercial value is huge. Blades are collectors that convert wind energy into electrical energy in wind turbines. Whether it has reliable performance is a key factor that the blade industry must pay attention to. The core technologies of wind turbine blades include aerodynamic airfoil design, load calculation, structural design verification, material selection, production and manufacturing, testing and other comprehensive integrated technologies. Among them, load calculation is a test of blade fatigue performance. In order to feedback the load distribution of the blade, it is necessary to install a grating fiber sensor on the blade to collect wavelength data and convert its corresponding relationship into a load value.

[0003] After installing the optical fiber load sensor at the root of the wind turbine blade, since the sensor cannot directly measure the blade load value, it is necessary to calibrate the initial reference wavelength value and sensitivity of the sensor through the blade gravity load before the real-time load value of the blade can be measured in real time. Its calibration method mainly includes the following steps: obtaining important static parameters of the wind turbine; obtaining real-time measurement data according to the specific calibration operation process; calculating the initial reference load and sensitivity coefficient of the load according to the calibration method provided in this article; writing the initial reference load and sensitivity coefficient into the wind turbine control system, collecting the calculated load value in real time according to the conversion formula, and checking the correctness and consistency of the loads of the three blades.

[0004] After a long period of operation, the loads of the three blades of the wind turbine will deviate to a certain extent. The main reason is that in the production of the sensor, due to its high sensitivity of optical fiber, it cannot be very stable. As time goes by, the initial reference wavelength of the optical fiber load sensor will shift with the increase of time, resulting in a large measurement error when measuring the blade load. Summary of the invention

[0005] The purpose of the present invention is to address the problem that the initial reference wavelength of the existing optical fiber load sensor will shift over time, resulting in a large measurement error when measuring the blade load. To address this problem, a real-time monitoring and correction method for the initial reference wavelength of the blade root load sensor of a wind turbine is proposed.

[0006] In order to achieve the above object, the present invention adopts the following technical solution:

[0007] The method for real-time monitoring and correction of the initial reference wavelength of a blade root load sensor of a power generation wind turbine comprises the following steps:

[0008] S1, mounting the grating optical fiber sensor on the blade, and calibrating the initial reference wavelength and sensitivity coefficient of the grating optical fiber sensor;

[0009] S2, collect the real-time wavelength of the grating optical fiber sensor;

[0010] S3, calculating the real-time load of the blade based on the real-time wavelength, the initial reference wavelength and the sensitivity coefficient;

[0011] S4, based on the load numerical logic, determining whether the initial reference wavelength of the grating optical fiber sensor needs to be corrected;

[0012] S5. If the initial reference wavelength of the grating optical fiber sensor needs to be corrected, the wavelength data of the grating optical fiber sensor is collected in real time;

[0013] According to λ i0_change =λ i0 -(λ i0_new -λ i0 ),

[0014] Among them, λ i0_change is the initial reference wavelength of each grating fiber sensor after modification; i0 is the reference wavelength of each grating fiber sensor before modification,

[0015] Correcting the initial reference wavelength of the grating optical fiber sensor;

[0016] S6. The fan master controller uses the initial reference wavelength and sensitivity coefficient of the new grating fiber sensor to calculate the real-time load.

[0017] As a further preferred embodiment of the present invention, S1 comprises the following steps:

[0018] S11, take the same section of the blade, take the center point of the leeward side as point P1, and install the grating fiber optic sensors at equal intervals of 90° clockwise;

[0019] S12, obtaining basic parameters of each blade of the wind turbine, wherein the basic parameters include: weight, center of mass position, nacelle elevation angle and blade cone angle;

[0020] S13, acquiring dynamic data of each blade of the wind turbine, wherein the dynamic data includes a dynamic real-time wavelength, a center wavelength, a pitch angle, and an azimuth angle;

[0021] S14. At low wind speed, combined with blade weight and fan parameters,

[0022] according to

[0023] Where m is the mass of the blade, g is the acceleration of gravity, r is the position of the center of gravity of the blade, C, T, P and A are the azimuth angle, impeller inclination angle, blade pitch angle and impeller azimuth angle respectively, and the theoretical flapping load M_flap and theoretical swing load M_edge are calculated;

[0024] S15. Based on the linear superposition principle, the multivariate linear regression method and normalization method are used to establish the relationship between the blade wavelength change and the load.

[0025]

[0026] Among them, K 11 , K 12 , K 13 , K 14 is the load factor of the four grating fiber sensors in the swinging direction; K 21 , K 22 , K 23 , K 24 is the load coefficient of the four grating fiber sensors in the swing direction; λ1, λ2, λ3, λ4 are the real-time wavelengths collected by the four grating fiber sensors; λ 10 , 20 , 30 , 40 is the initial reference wavelength of the four grating fiber sensors;

[0027] S16. Based on the least square method, the initial reference wavelength and sensitivity coefficient of each sensor are obtained, and the sensitivity coefficient is the load calibration coefficient.

[0028] As a further preference of the present invention, according to

[0029] Edge ni =K 11 *(λ 1i -λ 10 ) / λ 10 +K 12 *(λ 2i -λ 20 ) / λ 20 +K 13 *(λ 3i -λ 30 ) / λ 10 +K 14 *(λ 4i

[0030] -λ 40 ) / λ 40 ,

[0031] Flap ni =K 21 *(λ 1i -λ10 ) / λ 10 +K 22 *(λ 2i -λ 20 ) / λ 20 +K 23 *(λ 3i -λ 30 ) / λ 10 +K 24 *(λ 4i

[0032] -λ 40 ) / λ 40 ,

[0033] Where n is the blade number; i is the real-time wavelength number; λ 1i , 2i , 3i , 4i is the real-time wavelength of the four grating fiber sensors; Edge ni Flap is the real-time swing load of each blade; ni is the real-time flapping load of each blade.

[0034] As a further preferred embodiment of the present invention, S4 comprises the following steps:

[0035] S41, real-time monitoring of the load data of each blade of the wind turbine;

[0036] S42, calculate the real-time swing load Edge of the three blades within 10 minutes ni and real-time flap load ni The average value of

[0037] S43, according to

[0038] Per_Dev1=|m_edge 1i -m_edge 2i | / m_edge 1i ,

[0039] Per_Dev2=|m_edge 2i -m_edge 3i | / m_edge 2i ,

[0040] Per_Dev3=|m_edge 1i -m_edge 3i | / m_edge 3i ,

[0041] Among them, m_edge 1i 、m_edge2i 、m_edge 3i is the ten-minute average value of the three blade swing loads; m_flap 1i 、m_flap 2i 、m_flap 3i is the ten-minute average value of the flapping load of the three blades; Per_Dev1, Per_Dev2, and Per_Dev3 are the deviation percentages of the three blades respectively;

[0042] S44. When at least one of the deviation percentages of the three blades exceeds 10%, it is necessary to correct the initial reference wavelength of the grating optical fiber sensor.

[0043] As a further preferred embodiment of the present invention, S5 comprises the following steps:

[0044] S51, collecting wavelength data of a positive grating optical fiber sensor of a wind turbine running at a wind speed of less than 5 m / s, a pitch angle of three blades of 0, and a rotation speed of less than 0.5 to 2 rpm in real time; collecting the data in months, and performing calculations when the wavelength data is collected for a period of more than 5 hours;

[0045] S52, divide the wind speed into sections, select the peak and valley values ​​of the wind speed within 0.5m / s in each section, rotate one circle, and calculate the average value d ik , i represents the i-th sensor, k represents the k-th segment;

[0046] S53, according to λ i0_new =(d i1 +d i2 +d i3 +d i4 +d i5 ) / 5, the average value λ of the same grating fiber sensor in the whole wind speed range is calculated i0_new ;

[0047] S54, according to λ i0_change =λ i0 -(λ i0_new -λ i0 ),

[0048] Among them, λ i0 is the initial reference wavelength of each grating fiber sensor before modification, and the initial reference wavelength λ of each grating fiber sensor after modification is calculated. i0_change .

[0049] The method for real-time monitoring and correction of the initial reference wavelength of the blade root load sensor of the power generation wind turbine proposed in the present invention has the following beneficial effects compared with the prior art:

[0050] 1. The present invention can reduce the initial reference wavelength offset problem by correcting the initial reference wavelength of the grating optical fiber sensor, realize real-time correction, improve the load measurement accuracy, and reduce the time loss of repeated calibration. BRIEF DESCRIPTION OF THE DRAWINGS

[0051] Figure 1 This is a schematic diagram of the installation position of the grating optical fiber sensor;

[0052] Figure 2 It is a schematic diagram of the loads in the X and Y directions of the blade;

[0053] Figure 3 is a schematic diagram of blade azimuth and blade inclination;

[0054] Figure 4 It is a schematic diagram of the load calibration process; DETAILED DESCRIPTION

[0055] The present invention is described in detail below with reference to the accompanying drawings and specific embodiments.

[0056] The patent of this invention provides a real-time monitoring and correction of the initial reference wavelength of the sensor, which can realize real-time correction of the initial reference wavelength of the sensor, thereby increasing the accuracy of the load data and reducing the power generation loss caused by the calibration process of the sensor.

[0057] The method for real-time monitoring and correction of the initial reference wavelength of a blade root load sensor of a power generation wind turbine comprises the following steps:

[0058] S1. Mount the grating optical fiber sensor on the blade and calibrate the initial reference wavelength and sensitivity coefficient of the grating optical fiber sensor.

[0059] The following steps are involved:

[0060] S11. Take the same section of the blade, take the center of the leeward side as point P1, and install the grating fiber optic sensors at 90° intervals clockwise; the installation angle deviation of each sensor is within 5°; Figure 1 shown.

[0061] S12. Obtain basic parameters of each blade of the wind turbine, wherein the basic parameters include: weight, center of mass position, nacelle elevation angle, and blade cone angle.

[0062] S13. Acquire dynamic data of each blade of the wind turbine, wherein the dynamic data includes a dynamic real-time wavelength, a central wavelength, a pitch angle, and an azimuth angle.

[0063] The blade loads are mainly in the X and Y directions of the blade root, also known as flapping loads and shimmy loads.

[0064] S14. At low wind speed, combined with blade weight and fan parameters,

[0065] according to

[0066] Where m is the blade mass, g is the acceleration of gravity, r is the position of the blade center of gravity, C, T, P and A are the azimuth angle, impeller inclination angle, blade pitch angle and impeller azimuth angle, respectively. The theoretical flapping load M_flap and the theoretical swing load M_edge are calculated.

[0067] S15. Based on the linear superposition principle, the multivariate linear regression method and normalization method are used to establish the relationship between the blade wavelength change and the load.

[0068]

[0069] Among them, K 11 , K 12 , K 13 , K 14 is the load factor of the four grating fiber sensors in the swinging direction; K 21 , K 22 , K 23 , K 24 is the load coefficient of the four grating fiber sensors in the swing direction; λ1, λ2, λ3, λ4 are the real-time wavelengths collected by the four grating fiber sensors; λ 10 , 20 , 30 , 40 is the initial reference wavelength of the four grating fiber sensors.

[0070] S16. Based on the least square method, the initial reference wavelength and sensitivity coefficient of each sensor are obtained, and the sensitivity coefficient is the load calibration coefficient.

[0071] After calibration is completed, the initial reference wavelength and sensitivity coefficient need to be entered into the fan controller when the fan is stopped.

[0072] S2, collect the real-time wavelength of the grating fiber sensor.

[0073] S3. Calculate the real-time load of the blade based on the real-time wavelength, the initial reference wavelength and the sensitivity coefficient.

[0074] according to

[0075] Edge ni =K 11 *(λ 1i -λ 10 ) / λ 10 +K 12 *(λ 2i -λ 20 ) / λ 20 +K13 *(λ 3i -λ 30 ) / λ 10 +K 14 *(λ 4i

[0076] -λ 40 ) / λ 40 ,

[0077] Flap ni =K 21 *(λ 1i -λ 10 ) / λ 10 +K 22 *(λ 2i -λ 20 ) / λ 20 +K 23 *(λ 3i -λ 30 ) / λ 10 +K 24 *(λ 4i

[0078] -λ 40 ) / λ 40 ,

[0079] Where n is the blade number; i is the real-time wavelength number; λ 1i , 2i , 3i , 4i is the real-time wavelength of the four grating fiber sensors; Edge ni Flap is the real-time swing load of each blade; ni is the real-time flapping load of each blade.

[0080] S4. Based on the load numerical logic, determine whether the initial reference wavelength of the grating optical fiber sensor needs to be corrected.

[0081] The specific steps include:

[0082] S41. Monitor the load data of each blade of the wind turbine in real time.

[0083] S42, calculate the real-time swing load Edge of the three blades within 10 minutes ni and real-time flap load ni The average value of .

[0084] S43, according to

[0085] Per_Dev1=|m_edge 1i -m_edge 2i | / m_edge1i ,

[0086] Per_Dev2=|m_edge 2i -m_edge 3i | / m_edge 2i ,

[0087] Per_Dev3=|m_edge 1i -m_edge 3i | / m_edge 3i ,

[0088] Among them, m_edge 1i 、m_edge 2i 、m_edge 3i is the ten-minute average value of the three blade swing loads; m_flap 1i 、m_flap 2i 、m_flap 3i is the ten-minute average value of the flapping load of the three blades; Per_Dev1, Per_Dev2, and Per_Dev3 are the deviation percentages of the three blades respectively.

[0089] S44. When at least one of the deviation percentages of the three blades exceeds 10%, it is necessary to correct the initial reference wavelength of the grating optical fiber sensor.

[0090] S5. If the initial reference wavelength of the grating optical fiber sensor needs to be corrected, the wavelength data of the grating optical fiber sensor is collected in real time;

[0091] The following steps are involved:

[0092] S51. Real-time collection of wavelength data of a positive grating optical fiber sensor of a wind turbine running at a wind speed below 5 m / s, with the pitch angles of the three blades all being 0 and the rotation speed within the range of 0.5 to 2 rpm; collection is performed in months, and calculation is performed when the wavelength data is collected for a period of more than 5 hours.

[0093] S52, divide the wind speed into sections, select the peak and valley values ​​of the wind speed within 0.5m / s in each section, rotate one circle, and calculate the average value d ik , i represents the i-th sensor, and k represents the k-th segment.

[0094] For example: Select the peak and valley values ​​of a stable rotation of 2.5 to 3 m / s and find their average value d i1 ;

[0095] Select the peak and valley values ​​of a stable rotation of 3 to 3.5 m / s and calculate their average value d i2 ;

[0096] Select the peak and valley values ​​of a stable rotation of 3.5 to 4 m / s and calculate their average value d i3 ;

[0097] Select the peak and valley values ​​of a stable rotation of 4 to 4.5 m / s and calculate their average value d i4 ;

[0098] Select the peak and valley values ​​of a stable rotation of 4.5 to 5 m / s and calculate their average value d i5 .

[0099] S53, according to λ i0_new =(d i1 +d i2 +d i3 +d i4 +d i5 ) / 5, the average value λ of the same grating fiber sensor in the whole wind speed range is calculated i0_new .

[0100] S54, according to λ i0_change =λ i0 -(λ i0_new -λ i0 ),

[0101] Among them, λ i0 is the initial reference wavelength of each grating fiber sensor before modification, and the initial reference wavelength λ of each grating fiber sensor after modification is calculated. i0_change .

[0102] Correct the initial reference wavelength of the grating fiber sensor.

[0103] The corrected initial reference wavelength deviates greatly from the previous value. As a result, when calculating the load, the load may be suddenly updated or load jumps may occur within a short period of time after input. Therefore, it is not recommended to input the value when the fan is operating normally. The parameters can be input when the fan is stopped.

[0104] S6. The fan master controller uses the initial reference wavelength and sensitivity coefficient of the new grating fiber sensor to calculate the real-time load.

[0105] The above shows and describes the basic principles, main features and advantages of the present invention. Those skilled in the art should understand that the above embodiments do not limit the present invention in any form, and any technical solution obtained by equivalent replacement or equivalent transformation falls within the protection scope of the present invention.

Claims

1. A method for real-time monitoring and correction of the initial reference wavelength of a blade root load sensor of a power generation wind turbine, characterized in that: The following steps are involved: S1, mounting the grating optical fiber sensor on the blade, and calibrating the initial reference wavelength and sensitivity coefficient of the grating optical fiber sensor; S2, collect the real-time wavelength of the grating optical fiber sensor; S3, calculating the real-time load of the blade based on the real-time wavelength, the initial reference wavelength and the sensitivity coefficient; S4, based on the load numerical logic, determining whether the initial reference wavelength of the grating optical fiber sensor needs to be corrected; S5. If the initial reference wavelength of the grating optical fiber sensor needs to be corrected, the wavelength data of the grating optical fiber sensor is collected in real time; According to i0_change =λ i0 -(l i0_new -l i0 ), Among them, λ i0 is the initial reference wavelength of each grating fiber sensor before modification, and the initial reference wavelength λ of each grating fiber sensor after modification is calculated. i0_change , Correcting the initial reference wavelength of the grating optical fiber sensor; S6. The fan master controller uses the initial reference wavelength and sensitivity coefficient of the new grating fiber sensor to calculate the real-time load.

2. The method for real-time monitoring and correction of the initial reference wavelength of the blade root load sensor of a wind turbine according to claim 1 is characterized in that: The S1 comprises the following steps: S11, take the same section of the blade, take the center point of the leeward side as point P1, and install the grating fiber optic sensors at equal intervals of 90° clockwise; S12, obtaining basic parameters of each blade of the wind turbine, wherein the basic parameters include: weight, center of mass position, nacelle elevation angle and blade cone angle; S13, acquiring dynamic data of each blade of the wind turbine, wherein the dynamic data includes a dynamic real-time wavelength, a center wavelength, a pitch angle, and an azimuth angle; S14. At low wind speed, combined with blade weight and fan parameters, according to Where m is the mass of the blade, g is the acceleration of gravity, r is the position of the center of gravity of the blade, C, T, P and A are the azimuth angle, impeller inclination angle, blade pitch angle and impeller azimuth angle respectively, and the theoretical flapping load M_flap and theoretical swing load M_edge are calculated; S15. Based on the linear superposition principle, the multivariate linear regression method and normalization method are used to establish the relationship between the blade wavelength change and the load. Among them, K 11 , K 12 , K 13 , K 14 is the load factor of the four grating fiber sensors in the swinging direction; K 21 , K 22 , K 23 , K 24 is the load coefficient of the four grating fiber sensors in the swing direction; λ1, λ2, λ3, λ4 are the real-time wavelengths collected by the four grating fiber sensors; λ 10 , 20 , 30 , 40 is the initial reference wavelength of the four grating fiber sensors; S16. Based on the least square method, the initial reference wavelength and sensitivity coefficient of each sensor are obtained, and the sensitivity coefficient is the load calibration coefficient.

3. The method for real-time monitoring and correction of the initial reference wavelength of the blade root load sensor of a wind turbine according to claim 1 is characterized in that: according to Edge ni =K 11 *(l 1i -l 10 ) / min 10 +K 12 *(l 2i -l 20 ) / min 20 +K 13 *(l 3i -l 30 ) / min 10 +K 14 *(l 4i -l 40 ) / min 40 , Flap ni =K 21 *(l 1i -l 10 ) / min 10 +K 22 *(l 2i -l 20 ) / min 20 +K 23 *(l 3i -l 30 ) / min 10 +K 24 *(l 4i -l 40 ) / min 40 , Where n is the blade number; i is the real-time wavelength number; λ 1i , 2i , 3i , 4i is the real-time wavelength of the four grating fiber sensors; Edge ni Flap is the real-time swing load of each blade; ni is the real-time flapping load of each blade.

4. The method for real-time monitoring and correction of the initial reference wavelength of a blade root load sensor of a wind turbine according to claim 1, characterized in that: The S4 comprises the following steps: S41, real-time monitoring of the load data of each blade of the wind turbine; S42, calculate the real-time swing load Edge of the three blades within 10 minutes ni and real-time flap load ni The average value of S43, according to Per_Dev1=|m_edge 1i -m_edge 2i | / m_edge 1i , Per_Dev2=|m_edge 2i -m_edge 3i | / m_edge 2i , Per_Dev3=|m_edge 1i -m_edge 3i | / m_edge 3i , Among them, m_edge 1i 、m_edge 2i 、m_edge 3i is the ten-minute average value of the three blade swing loads; m_flap 1i 、m_flap 2i 、m_flap 3i is the ten-minute average value of the flapping load of the three blades; Per_Dev1, Per_Dev2, and Per_Dev3 are the deviation percentages of the three blades respectively; S44. When at least one of the deviation percentages of the three blades exceeds 10%, it is necessary to correct the initial reference wavelength of the grating optical fiber sensor.

5. The method for real-time monitoring and correction of the initial reference wavelength of the blade root load sensor of a wind turbine according to claim 1, characterized in that: The S5 comprises the following steps: S51, collecting wavelength data of a positive grating optical fiber sensor of a wind turbine running at a wind speed of less than 5 m / s, a pitch angle of three blades of 0, and a rotation speed of less than 0.5 to 2 rpm in real time; collecting the data in months, and performing calculations when the wavelength data is collected for a period of more than 5 hours; S52, divide the wind speed into sections, select the peak and valley values ​​of the wind speed within 0.5m / s in each section, rotate one circle, and calculate the average value d ik , i represents the i-th sensor, k represents the k-th segment; S53, according to λ i0_new =(d i1 +d i2 +d i3 +d i4 +d i5 ) / 5, Calculate the average value λ of the same grating fiber sensor in the full wind speed range i0_new ; S54、according to no i0_change =λ i0 -(l i0_new -l i0 ), Among them, λ i0 is the initial reference wavelength of each grating fiber sensor before modification, and the initial reference wavelength λ of each grating fiber sensor after modification is calculated. i0_change .

Images