A calibration method for optical fiber sensors in wind turbine blades

Abstract

The present invention belongs to the technical field of wind power generation, and specifically relates to a calibration method for an optical fiber sensor in a wind power blade, including: S1, the blade is vertically cleared upward, a strain value is obtained, and an influence coefficient t of wind speed and strain value is obtained; S2, clearing again, the blade is vertically downward, a strain value is obtained, the strain value is compared with the strain value in S1, and a difference value k is obtained; S3, the blade is vertically upward, the current wind speed is obtained, the current wind speed is compared with the wind speed in S1, and the wind speed change Δx1 is obtained, and basic data f is obtained through the wind speed change Δx1, the influence coefficient t and the difference value k and assigned to the optical fiber sensor; the blade posture is subjected to less force, the influence of the blade's own deformation is reduced, the error magnitude is small, and the difference caused by the wind speed can be combined with the difference in wind speed and strain value to judge the difference, which can reduce the influence of the wind speed. The assigned data has a lower deviation from the theoretical data and is closer to the real data.

Description

Technical Field

[0001] The invention belongs to the technical field of wind power generation, and in particular relates to a calibration method for an optical fiber sensor in a wind power blade. Background Art

[0002] Due to the ever-changing environment faced by wind turbine blades in wind farms, coupled with the current industry's pursuit of lightweight wind turbine blades, the safety margin of wind turbine blades is getting lower and lower. In order to ensure the reliability of wind turbine blade operation, it is necessary to install fiber optic sensors at the corresponding positions on the wind turbine blades to monitor the load of wind turbine blades in real time. If monitoring is performed directly through fiber optic sensors, there will be deviations between the monitoring data and the actual operation of wind turbine blades. Therefore, it is necessary to pre-calibrate the fiber optic sensor so that the acquired monitoring data is closer to the real data.

[0003] The traditional calibration method is to first adjust the angle of the target blade to Figure 1 The parameters of the optical fiber sensor in the target blade are read and recorded, and then the same blade is rotated 180°, as shown in Figure 2 The parameters of the fiber optic sensor in the blade are read again in the posture shown, and the average value of the two parameters is calculated. The average value is used as the basic data of the blade and assigned to the fiber optic sensor. However, this method has the following problems: the load magnitude of the blade cross section is large in the horizontal posture, and the initial value is set in this posture, and the deviation of the read data is large. There is an elevation angle θ1 between the centerline of the cabin and the horizontal line, and there is also a cone angle θ2 between the centerline of the blade and the vertical line of the cabin, which makes the 0° posture of the blade different from the 180° posture. Therefore, the calibration data contains the differences caused by the elevation angle θ1 and the cone angle θ2, and the influence of wind speed on the data is not eliminated, resulting in a large magnitude of data error. Summary of the invention

[0004] The technical problem to be solved by the present invention is to provide a calibration method for optical fiber sensors in wind turbine blades, which can reduce the errors caused by the elevation angle θ1, the cone angle θ2 and the wind speed and reduce the errors in monitoring data.

[0005] Content of the present invention comprises the following steps:

[0006] S1. Adjust the blade to be vertically upward, and clear the optical fiber sensor on the blade. After clearing, obtain the strain value of the optical fiber sensor at the current wind speed, and obtain the influence coefficient t of the wind speed and the strain value under the posture by comparing the wind speed and the strain value;

[0007] S2, based on the blade posture in S1, the optical fiber sensor on the blade is reset again, and after reset, the blade posture is adjusted to be vertically downward, and the strain value of the optical fiber sensor is obtained, and the strain value is compared with the strain value in S1 to obtain a difference value k;

[0008] S3, adjust the blade posture to vertically upward, obtain the current wind speed, compare the current wind speed with the wind speed in S1, obtain the wind speed change Δx1, and obtain the basic data f through the wind speed change Δx1, influence coefficient t and difference value k, f=0.5k+t·Δx1; assign the basic data f to the optical fiber sensor.

[0009] Furthermore, after S3, the step further includes:

[0010] S4. Adjust the blade to a horizontal posture. In the horizontal posture, rotate the blade along its own circumferential direction. Read the actual strain value of the optical fiber sensor at different angles in the horizontal posture. Obtain the actual strain change based on the actual strain value of the optical fiber sensor at the same position at different angles in the horizontal posture. Fit the actual strain change with the theoretical strain change, and correct the parameters of the optical fiber sensor after assignment.

[0011] Furthermore, in S4, the fitting of the actual strain variation and the theoretical strain variation is a comparison between the actual strain variation and the theoretical strain variation, and a proportional coefficient t1 is obtained through the comparison, and the parameters of the assigned optical fiber sensor are corrected through the proportional coefficient t1.

[0012] Furthermore, in S4, the actual strain change is the actual strain change of a single optical fiber sensor, the theoretical strain change is the theoretical strain change of a single optical fiber sensor, and the correction of the optical fiber sensor parameters after assignment is a correction performed one by one according to the proportional coefficient t1 corresponding to each optical fiber sensor.

[0013] Furthermore, in S4, when obtaining the actual strain value of the optical fiber sensor on the blade, the different angles of the horizontal posture of the blade are 0°, 90°, 180°, and 270° of the blade 1 itself.

[0014] Furthermore, in S1, the comparison between the wind speed and the strain value is a comparison between a change Δx of the wind speed per unit time and a change Δy of the strain value per unit time, and the influence coefficient t=Δy / Δx.

[0015] Furthermore, the comparison between the strain value in S2 and the strain value in S1 is the difference between the average value of the strain value per unit time and the average value of the strain value per unit time in S1.

[0016] Furthermore, in S3, the current wind speed is obtained and compared with the wind speed in S1, so as to obtain the wind speed in the current unit time, and compare the average value of the wind speed in the unit time with the average value of the wind speed in the unit time in S1, and the wind speed change Δx1 is the difference between the two wind speed average values.

[0017] Furthermore, S2 also includes that if the wind speed in the unit time when the strain value is obtained is different from the wind speed in the unit time in S1, the strain value in S2 is corrected by the influence coefficient t obtained in S1 according to the difference variable between the two wind speeds.

[0018] Furthermore, the wind speed is less than 5m / s.

[0019] The beneficial effects of the present invention are that the blade posture is subjected to less force, the influence of the blade's own deformation is reduced, the error magnitude is small, and the error is lower than that calibrated in the horizontal posture; the error influence caused by the elevation angle θ1 and the cone angle θ2 is reduced, and the error is lower.

[0020] The wind speed can be combined with the strain value difference of each optical fiber sensor under different wind speeds in the same posture to determine the difference caused by wind speed. Compared with the previous calibration method, the influence of wind speed can be reduced. Since the strain of the blade is the smallest when it is vertically upward, after excluding the influence of wind speed, 1 / 2 of the difference k between vertically upward and vertically downward is assigned to the blade in the vertically upward posture. At this time, the deviation between the assigned data of the blade in the vertically upward posture and the theoretical data is the strain value of the blade in the vertically upward posture. This assigned data is used as the reference data of the blade, which has lower error and is closer to the real data. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] Attached Figure 1 This is a schematic diagram of the first posture of the blade.

[0022] Attached Figure 2 This is a schematic diagram of the second posture of the blade.

[0023] Attached Figure 3 This is a schematic diagram of the third posture of the blade.

[0024] Attached Figure 4 This is a schematic diagram of the fourth posture of the blade.

[0025] Attached Figure 5 Schematic diagram of cone angle and elevation angle.

[0026] Attached Figure 6 for Figure 1 Schematic diagram of the blade root twist angle in the prior art calibration method under the AA cross-sectional view.

[0027] Attached Figure 7 for Figure 1 Schematic diagram of the blade root in the calibration method of the present invention in the AA cross-sectional view.

[0028] In the figure, 1-blade; 2-optical fiber sensor; 3-nacelle. DETAILED DESCRIPTION

[0029] The present invention provides a calibration method for an optical fiber sensor in a wind turbine blade, the calibration method comprising the following steps:

[0030] S1. Adjust the posture of blade 1 to be vertically upward, as shown in Figure 3 As shown in the blade posture, after the blade 1 is in a vertical upward posture without shaking, the optical fiber sensor 2 on the blade 1 is cleared, and the strain value of the optical fiber sensor 2 is obtained at the current wind speed after clearing, and the influence coefficient t of the wind speed and the strain value under the posture is obtained by comparing the wind speed with the strain value; wherein, the comparison of the wind speed and the strain value is the comparison of the wind speed change Δx per unit time with the strain value change Δy per unit time, for example, when the wind speed change per unit time is Δx, the corresponding strain value change per unit time is Δy, and the wind speed influence coefficient on the strain value t=Δy / Δx.

[0031] S2. Based on the posture of blade 1 in S1, the optical fiber sensor 2 on blade 1 is reset again. After reset, the posture of blade 1 is adjusted to be vertically downward, that is, Figure 4 The blade posture is shown, and the strain value of the optical fiber sensor 2 is obtained under this posture, and the strain value is compared with the strain value in S1, specifically, the difference between the average value of the strain value per unit time and the average value of the strain value per unit time in S1, so as to obtain the difference value k; wherein, in S2, if the average value of the wind speed per unit time when the strain value is obtained is different from the average value of the wind speed per unit time in S1, then according to the difference variable between the two wind speeds, the strain value in S2 is corrected by the influence coefficient t obtained by S1 to obtain a more realistic strain value, so as to eliminate the influence of different wind speeds on the difference value k.

[0032] S3, adjust the attitude of blade 1 to vertically upward again, obtain the wind speed in the current unit time, compare the current wind speed with the wind speed in S1, specifically compare the average value of the wind speed in the current unit time with the average value of the wind speed in S1, obtain the wind speed change Δx1, that is, the difference between the two wind speed averages, obtain the basic data f through the wind speed change Δx1, the influence coefficient t and the difference value k, f=0.5k+t·Δx1; assign the basic data f to the optical fiber sensor 2. Among them, the wind speeds described in S1, S2, and S3 are all less than 5m / s, and the unit time is preferably 2min.

[0033] In the present invention, each optical fiber sensor 2 will obtain its corresponding basic data f through the above steps, and the assignment referred to in S3 is to assign the corresponding basic data f to each optical fiber sensor 2.

[0034] In such Figure 1 and Figure 2In the horizontal posture, the main load-bearing parts of the blade 1 are the core and the blade root area, and the core and the blade root area are the installation areas of the optical fiber sensor 2, and the strain magnitude is large. Figure 3 In the vertical upward posture, the main load-bearing component of the blade is the beam, the core and the root area are subjected to the smallest force, and the strain magnitude is the smallest. At this time, all the optical fiber sensors 2 on the blade 1 are reset to zero, and the error is lower than that in other postures, and the error caused by the deformation of the optical fiber sensor 2 itself during the construction process can be eliminated at the same time.

[0035] like Figure 5 As shown, there is an elevation angle θ1 between the center line of the cabin 3 and the horizontal line, and there is a cone angle θ2 between the vertical line of the cabin 3 and the center line of the blade 1. The present invention makes the blade 1 vertically upward, clears the optical fiber sensor 2, and then makes the blade 1 vertically downward and reads the data of the optical fiber sensor 2. At this time, the difference in the read data only comes from the influence of wind speed and the elevation angle θ1 and the cone angle θ2. When vertically downward, the influencing factor θ of blade 1 is down = The deviation of the elevation angle θ1 plus the cone angle θ2 and the effect of wind speed (θ down =θ1+θ2), wind speed influence; when vertically upward, the influencing factor of blade 1 is θ up = The deviation of the elevation angle θ1 minus the cone angle θ2 (θ up =θ1-θ2), wind speed influence; after eliminating the influence of wind speed, the difference between blade 1 vertically upward and vertically downward is: θ=(θ1+θ2)-(θ1-θ2))=2·θ2, that is, the difference is only caused by the cone angle θ2.

[0036] The calibration method adopted by the present invention can be combined with the wind speed condition, and by comparing the strain value difference of each optical fiber sensor 2 at different wind speeds in the same posture, the difference caused by the wind speed can be determined, and compared with the previous calibration method, the influence of the wind speed can be reduced. Since the strain of the blade 1 in the vertical upward direction is the smallest, after eliminating the influence of the wind speed, 1 / 2 of the difference value k between the vertical upward direction and the vertical downward direction is assigned to the vertical upward posture of the blade 1. At this time, the deviation between the assigned data of the blade 1 in the vertical upward direction and the theoretical data is the strain value of the blade 1 in the vertical upward direction. The assigned data is used as the reference data of the blade 1, which has a lower error and is closer to the real data.

[0037] The calibration method provided by the present invention is compared with Figure 1 and 2 In the traditional calibration method shown, the posture of blade 1 is subjected to less force, the influence of the deformation of blade 1 itself is reduced, the error magnitude is small, and the error is lower than that calibrated in the horizontal posture; the error influence caused by the elevation angle θ1 and the cone angle θ2 is reduced, and the error is lower.

[0038] The calibration method provided by the present invention, after assigning a value to each optical fiber sensor 2 in S3, further comprises:

[0039] S4, adjust blade 1 to Figure 1 The horizontal posture shown, on the basis of maintaining the horizontal posture, the blade 1 is rotated along its own circumferential direction, and the actual strain value of the optical fiber sensor 2 is read at different angles of the horizontal posture, and the actual strain change is obtained according to the actual strain value of the optical fiber sensor 2 at different angles of the horizontal posture at the same position, and the theoretical strain change value is obtained through the result of finite element calculation, and the actual strain change is fitted with the theoretical strain change, and the parameters of the optical fiber sensor 2 after the assignment are corrected, so that the monitoring data is closer to the real data. Among them, when obtaining the actual strain value of the optical fiber sensor 2 on the blade 1, the different angles of the horizontal posture of the blade 1 are 0°, 90°, 180°, and 270° of the blade 1 itself.

[0040] Since there is a certain twist angle θ3 between blade 1 and the main engine, such as Figure 6 As shown in the figure, (a) is a schematic diagram of the blade root torsion angle at 0°, (b) is a schematic diagram of the blade root torsion angle at 90°, (c) is a schematic diagram of the blade root torsion angle at 180°, and (d) is a schematic diagram of the blade root torsion angle at 270°. The existence of the above-mentioned torsion angle θ3 makes there be a difference of torsion angle θ3 between the pitch zero degree and the zero degree of blade 1 itself, while the traditional horizontal calibration method uses the pitch angle of the main engine as a reference. Since the strain magnitude of the blade is large under the horizontal posture, the error magnitude caused by the blade root torsion angle is large. The present invention adopts 0°, 90°, 180°, and 270° of blade 1 itself, and takes 0° of blade 1 itself as the reference. Figure 7 As shown, (a) is a schematic diagram of the blade root of the blade 1 itself when it is 0°, (b) is a schematic diagram of the blade root of the blade 1 itself when it is 90°, (c) is a schematic diagram of the blade root of the blade 1 itself when it is 180°, and (d) is a schematic diagram of the blade root of the blade 1 itself when it is 270°. Since the present invention uses the angle of the blade 1 itself as a reference, the error effect of the torsion angle θ3 on the data during the calibration process can be reduced.

[0041] Among them, in the present invention, if Figure 7 As shown, the optical fiber sensors 2 in the root area of ​​the blade 1 are circumferentially distributed in the inner cavity of the root area. The optical fiber sensors 2 are specifically divided into four groups, corresponding to 0°, 90°, 180°, and 270° of the blade 1 itself.

[0042] In S4, the fitting of the actual strain variation and the theoretical strain variation is specifically as follows: the actual strain variation is compared with the theoretical strain variation, and the proportional coefficient t1 is obtained through the comparison. For example, the theoretical strain value of the optical fiber sensor 1 at the blade root position in the horizontal posture of 0° is X1, and the theoretical strain value in the horizontal posture of 180° is X2, then the theoretical strain variation is X2-X1; during measurement, the actual strain value of the optical fiber sensor 1 at the blade root position in the horizontal posture of 0° is Y1, and the actual strain value in the horizontal posture of 180° is Y2, then the actual strain variation is Y2-Y1; by comparing the actual strain variation with the theoretical strain variation, the proportional coefficient t1 is obtained, t1=(Y2-Y1) / (X2-X1), the proportional coefficient is the influence of different angles of the same posture on the data, the proportional coefficient t1 is multiplied by the original set parameters of the optical fiber sensor 2, so as to correct the parameters of the optical fiber sensor 2 after the assignment.

[0043] Among them, the actual strain change is the actual strain change of a single optical fiber sensor 2, and the theoretical strain change is the theoretical strain change of a single optical fiber sensor 2. The correction of the parameters of the optical fiber sensor 2 after assignment is a one-by-one correction of the proportional coefficient t1 corresponding to each optical fiber sensor 2 obtained according to S4.

[0044] Compared with the traditional horizontal calibration method, the calibration method proposed in the present invention compares the actual data with the theoretical data during calibration, obtains the corresponding proportional relationship through the deviation between the actual data and the theoretical data, makes the overall data trend consistent with the theoretical result trend, and corrects the optical fiber sensor 2 one by one after the assignment through the proportional relationship, so that the monitoring data is closer to the real data, and the monitoring of the blade 1 is more effective. At the same time, the calibration method of the present invention has more calibrated angles, and the strain difference of the optical fiber sensor 2 at the same position of the blade 1 under different postures can be compared, and the strain difference of the optical fiber sensor 2 at different positions under the same posture can also be compared.

Claims

1. A method for calibrating an optical fiber sensor in a wind turbine blade, characterized in that: The following steps are involved: S1, adjusting the posture of the blade (1) to be vertically upward, and clearing the optical fiber sensor (2) on the blade (1), obtaining the strain value of the optical fiber sensor (2) at the current wind speed after clearing, and obtaining the influence coefficient t of the wind speed and the strain value at the posture by comparing the wind speed and the strain value; S2, based on the posture of the blade (1) in S1, the optical fiber sensor (2) on the blade (1) is reset again, and after reset, the posture of the blade (1) is adjusted to be vertically downward, and the strain value of the optical fiber sensor (2) is obtained, and the strain value is compared with the strain value in S1 to obtain a difference value k; S3, adjusting the attitude of the blade (1) to be vertically upward again, obtaining the current wind speed, comparing the current wind speed with the wind speed in S1, obtaining the wind speed change Δx1, obtaining basic data f through the wind speed change Δx1, the influence coefficient t and the difference value k, f=0.5k+t·Δx1; assigning the basic data f to the optical fiber sensor (2); S4, adjusting the blade (1) to a horizontal posture, rotating the blade (1) along its own circumferential direction in the horizontal posture, reading the actual strain value of the optical fiber sensor (2) at different angles in the horizontal posture, and obtaining the actual strain variation according to the actual strain value of the optical fiber sensor (2) at the same position at different angles in the horizontal posture, fitting the actual strain variation with the theoretical strain variation, and correcting the parameters of the optical fiber sensor (2) after the assignment; In S1, the comparison between wind speed and strain value is the comparison between the wind speed change Δx per unit time and the strain value change Δy per unit time, and the influence coefficient t=Δy / Δx; In S4, the fitting of the actual strain variation and the theoretical strain variation is a comparison between the actual strain variation and the theoretical strain variation, a proportional coefficient t1 is obtained through the comparison, and the parameters of the assigned optical fiber sensor (2) are corrected through the proportional coefficient t1; In S4, the actual strain change is the actual strain change of a single optical fiber sensor (2), the theoretical strain change is the theoretical strain change of a single optical fiber sensor (2), and the correction of the parameters of the optical fiber sensor (2) after the assignment is performed one by one according to the proportional coefficient t1 corresponding to each optical fiber sensor (2).

2. The calibration method of the optical fiber sensor in the wind turbine blade according to claim 1, characterized in that: In S4, when obtaining the actual strain value of the optical fiber sensor (2) on the blade (1), the different angles of the horizontal posture of the blade (1) are 0°, 90°, 180°, and 270° of the blade (1) itself.

3. The calibration method of the optical fiber sensor in the wind turbine blade according to claim 1, characterized in that: The comparison between the strain value in S2 and the strain value in S1 is the difference between the average value of the strain value per unit time and the average value of the strain value per unit time in S1.

4. The calibration method of the optical fiber sensor in the wind turbine blade according to claim 1, characterized in that: In S3, the current wind speed is obtained and compared with the wind speed in S1, so as to obtain the wind speed in the current unit time, and compare the average value of the wind speed in the unit time with the average value of the wind speed in S1, and the wind speed change Δx1 is the difference between the two wind speed average values.

5. The calibration method of the optical fiber sensor in the wind turbine blade according to claim 3 or 4, characterized in that: S2 also includes that if the wind speed in the unit time when the strain value is obtained is different from the wind speed in the unit time in S1, the strain value in S2 is corrected according to the difference variable between the two wind speeds through the influence coefficient t obtained in S1.

6. The calibration method of the optical fiber sensor in the wind turbine blade according to any one of claims 1 to 4, characterized in that: The wind speeds are all less than 5m / s.

Images