Fan blade ice detection method based on FBG sensor bowstring mechanical response

Abstract

The application relates to the technical field of fan blade ice detection, and discloses a fan blade ice detection method based on FBG sensor bowstring mechanical response, which comprises the following steps: first step: after a substrate FBG strain sensor is pre-tensioned, the substrate FBG strain sensor is pasted on the leading edge of a fan blade to obtain a bowstring structure composed of the sensor and the fan blade; second step: a substrate FBG strain sensor signal curve of the fan blade under different working conditions and without icing is obtained in one complete cycle by an optical fiber grating demodulator, and a characteristic mark is made; third step: a substrate FBG strain sensor strain signal curve in one cycle of the fan blade is obtained by the optical fiber grating demodulator; and fourth step: the signal curve of the substrate FBG strain sensor without icing is compared to obtain the 'peak clipping' condition of the current strain signal curve trough, and the current icing condition is judged.

Description

Technical Field

[0001] This invention relates to the field of wind turbine blade ice detection technology, specifically a wind turbine blade ice detection method based on the bowstring mechanical response of an FBG sensor. Background Technology

[0002] With the implementation of my country's sustainable development strategy, wind power, as a clean and green renewable energy source, has developed rapidly, specifically manifested in a significant increase in installed capacity. However, wind power is sensitive to weather conditions and exhibits a certain degree of instability. In humid regions during winter and spring, supercooled water droplets with temperatures below 0°C collide with the blade surface. Smaller droplets condense directly into frost, while larger droplets form a thin water film that flows under the combined influence of wind, centrifugal force, and gravity, eventually condensing into clear ice. Whether it's frost, clear ice, or a mixture of both, all of these alter the aerodynamic performance of the blades, increasing their load. This not only reduces the power generation of the wind turbine but also significantly shortens the blade's lifespan, which is particularly severe for large-capacity wind turbines. Therefore, effective icing detection measures are necessary. Due to the remote installation locations of wind turbines, the similarity in color between icing and blades, and the risk of ice detachment, visual inspection alone is insufficient to effectively identify the icing status. Current wind turbine blade icing detection methods can be divided into two categories: direct detection methods and indirect detection methods. Direct detection methods directly measure changes in the mass, conductivity, inductance, reflection characteristics, and capacitance of the blades before and after icing. Indirect detection methods rely on changes in current wind speed, temperature, humidity, and the generator's power output and resistance. Currently popular icing detection methods include vibration frequency variation, icing sensor methods, image recognition methods, and neural network methods. Vibration frequency variation methods use piezoelectric vibration sensors to measure changes in vibration frequency before and after icing. Icing sensor methods measure the intensity of light reflected back from iced objects after infrared light is irradiated; different ice thicknesses have different reflectivities, making this the most widely used method. Image recognition methods use cameras to capture images and machine vision technology to identify the presence or absence of icing. Neural network methods utilize the large amounts of data generated by supervisory control and data acquisition (SCADA) systems widely used in the wind power industry, employing data-driven methods for icing identification. Although these methods exhibit varying degrees of sensitivity, practical applications still face challenges, one of which is the multi-channel and interference issues in electrical signal transmission.

[0003] When wind turbines operate in different geographical and meteorological environments, such as different wind and sand conditions, temperature and humidity, and wind fields, the vibration frequency and electrical properties of the blades will also be different. Therefore, the existing technologies mentioned above, such as the vibration frequency change method and image recognition method, need to be debugged according to the specific wind power generation site, which is quite difficult to apply.

[0004] When sensors are installed on large blades, multiple sensors are usually required. Schemes that use electrical signals for transmission are prone to problems such as a large number of signal channels, signal attenuation, electromagnetic interference between signal lines, and vulnerability to lightning strikes.

[0005] Existing technologies, such as the icing sensor method, require mounting on the blade surface. However, the sensor itself has a large thickness, which can easily affect the aerodynamic shape of the blade. To address this, we propose a wind turbine blade icing detection method based on the bowstring mechanical response of an FBG sensor. Summary of the Invention

[0006] (a) Technical problems to be solved

[0007] To address the shortcomings of existing technologies, this invention provides a method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor, thus solving the aforementioned problems.

[0008] (II) Technical Solution

[0009] To achieve the above-mentioned objectives, the present invention provides the following technical solution: a method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor, comprising the following steps:

[0010] Step 1: After pre-tensioning the substrate-type FBG strain sensor, it is pasted onto the leading edge of the wind turbine blade to obtain a bowstring structure composed of the sensor and the wind turbine blade;

[0011] Step 2: Obtain the substrate FBG strain sensor signal curves of a complete cycle of wind turbine blade operation under different working conditions without icing using a fiber optic grating demodulator, and mark the features accordingly;

[0012] Step 3: Obtain the strain signal curve of the substrate FBG strain sensor for one cycle of the wind turbine blade using a fiber optic grating demodulator.

[0013] Step 4: Compare the signal curves of the substrate FBG strain sensor when it is not covered with ice to obtain the "peak clipping" of the current strain signal curve and determine the current icing situation.

[0014] Preferably, the substrate-type FBG strain sensor in the first step consists of a "π"-shaped substrate and an optical fiber encapsulated in the substrate. The substrate is a glass fiber sheet with a thin middle section and thick ends. When the substrate is pasted onto the blade surface, there is a gap h between the thinner middle section of the substrate and the blade. The size of the gap h affects the ice detection sensitivity of the sensor. The thicker ends are the pasting sections, which are fixed to the blade by adhesive.

[0015] The preferred method for obtaining the signal curve of the substrate-type FBG strain sensor is as follows:

[0016] S1: When the blade is not bent or deformed, the sensor is in a tensile state. At this time, the strain value output by the substrate FBG strain sensor is recorded as ε0.

[0017] S2: When the blade swings down, the blade surface undergoes bending and tensile deformation. The sensor is further tightened, the strain increases, and the gap between the sensor and the blade decreases until the sensor contacts the blade. At this time, the strain value of the substrate FBG strain sensor is recorded as ε1. Therefore, relative to when the blade is not bent, the strain change value of the substrate FBG strain sensor is Δε1=ε1-ε0.

[0018] S3: The blade swings upward, and bending deformation and compressive strain occur on the upper surface of the blade. The sensor relaxes, the strain decreases, and the gap between the substrate FBG strain sensor and the blade increases. At this time, the strain value of the substrate FBG strain sensor is recorded as ε2. Therefore, relative to when the blade is not bent, the strain change value of the substrate FBG strain sensor is Δε2=ε0-ε2.

[0019] Preferably, Δε1>Δε2 and Δε2>0.

[0020] Preferably, when icing occurs, the ice accumulation in the gap will stick the sensor to the blade. At this time, the sensor is in close contact with the blade, and the strain detected by the FBG sensor when it swings up and down is the actual strain on the blade surface. At this time, Δε1=Δε2, and the sensor output signal is a standard sine signal.

[0021] (III) Beneficial Effects

[0022] Compared with the prior art, the present invention provides a method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor, which has the following advantages:

[0023] 1. This wind turbine blade icing detection method based on the bowstring mechanical response of an FBG sensor uses the flapping and bending of the blade during operation as the excitation signal. When the blade is not iced, the FBG sensor on the bowstring structure outputs a peak-shaving signal. When icing occurs, the accumulated ice adheres to the gaps in the bowstring structure, causing the peak-shaving phenomenon in the FBG sensor output signal to disappear. The presence or absence of icing can be determined by comparing the presence or absence of signal peaks. Since the bowstring structure is located on the blade surface, icing directly affects the bowstring structure. Therefore, this method is less affected by external environmental interference, is suitable for icing detection in various environments, and has strong versatility.

[0024] 2. The wind turbine blade ice detection method based on the bowstring mechanical response of FBG sensor uses fiber Bragg grating (FBG) for ice detection. Since FBG has the characteristics of anti-electromagnetic interference, high accuracy, and low signal attenuation, and allows multiple sensors to be connected in series to form a quasi-distributed optical fiber, it can solve the problems of multiple signal channels and interference in electrical signal measurement.

[0025] 3. The wind turbine blade ice detection method based on the bowstring mechanical response of the FBG sensor has a small thickness. When the sensor is attached to the blade surface for ice detection, it hardly changes the aerodynamic shape of the blade and does not affect the blade life. Attached Figure Description

[0026] Figure 1 This is a schematic diagram of the FBG sensor structure of the present invention;

[0027] Figure 2 This is a schematic diagram of the bowstring structure of the present invention;

[0028] Figure 3 This is a schematic diagram of the deformation of the bowstring structure of the present invention;

[0029] Figure 4 This is a schematic diagram of the sensor's output signal. Detailed Implementation

[0030] The technical solutions of the embodiments of the present invention will be clearly and completely described below with reference to the accompanying drawings. Obviously, the described embodiments are only some embodiments of the present invention, and not all embodiments. Based on the embodiments of the present invention, all other embodiments obtained by those skilled in the art without creative effort are within the scope of protection of the present invention.

[0031] Please see Figure 1-4 A method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor, comprising the following steps:

[0032] Step 1: After pre-tensioning the substrate-type FBG strain sensor, attach it to the leading edge of the wind turbine blade to obtain a bowstring structure composed of the sensor and the wind turbine blade.

[0033] The substrate-type FBG strain sensor consists of a "π"-shaped substrate and an optical fiber encapsulated within the substrate, such as... Figure 1 As shown. The "π"-shaped substrate has a cross-section made of glass fiber sheet that is thin in the middle and thick at both ends. When the substrate is bonded to the blade surface, there is a gap h between the thinner middle section of the substrate and the blade. The size of the gap h affects the ice detection sensitivity of the sensor. The thicker ends are the bonding sections, which are fixed to the blade with adhesive.

[0034] After being pre-tensioned, the sensor is attached to the leading edge of the blade. Since icing on wind turbine blades first occurs at the blade tip leading edge, installing the sensor here helps in the early identification of icing, enabling timely detection and handling. When the blade is subjected to wind load during operation, it flaps and bends. The taut sensor can be considered as a "string," and the wind turbine blade as a "bow." The FBG sensor and the wind turbine blade form a bowstring structure, as shown below. Figure 2 As shown, with the formation of ice, the ice fills the gap between the sensor and the blade, causing them to stick together. The deformation characteristics of the sensor are different before and after icing, and its output strain signal is also different.

[0035] Step 2: Obtain the FBG strain sensor signal curves of a complete cycle of wind turbine blade operation under different operating conditions without icing using a fiber optic grating demodulator, and mark the features.

[0036] The strain signal curves of the wind turbine blades under different operating conditions (un-iced) were obtained using a fiber Bragg grating demodulator. These different operating conditions included varying wind speeds, blade angles of attack, and temperatures. The aim was to reduce interference from these operating conditions, obtain strain signal curves under multiple conditions, and improve the accuracy of icing detection.

[0037] Step 3: Obtain the FBG strain sensor signal curve for one complete cycle of wind turbine blade operation.

[0038] The strain signal curve of the wind turbine blade within one cycle is obtained using a fiber Bragg grating demodulator. Since judging the icing condition requires comparing the strain signals within one cycle, the minimum icing detection interval of this method is one blade rotation cycle, which is related to the blade rotation speed.

[0039] Step 4: Compare the signal curves of the FBG strain sensor when it is not covered with ice to obtain the "peak clipping" of the current strain signal curve trough and determine the current icing situation.

[0040] The specific detection principle is as follows:

[0041] After being pre-tensioned, the FBG sensor is attached to the blade. When the blade is not bent or deformed, the sensor is in a tensile state, and the strain value output by the sensor is recorded as ε0. When the blade bends (in this example, the sensor is attached to the upper surface of the wind turbine blade), depending on the bending direction of the blade, the sensor undergoes two different deformation processes, such as... Figure 3 As shown:

[0042] (1) As the blade swings downward, bending and tensile deformations occur on the blade surface. The sensor is further tightened, the strain increases, and the gap between the sensor and the blade decreases until the sensor contacts the blade. At this time, the strain value of the sensor is recorded as ε1. Therefore, relative to when the blade is not bent, the strain change of the sensor is Δε1=ε1-ε0.

[0043] (2) When the blade swings up, the upper surface of the blade will bend and compressive strain. The sensor relaxes, the strain decreases, and the gap between the sensor and the blade increases. At this time, the strain value of the sensor is recorded as ε2. Therefore, relative to when the blade is not bent, the strain change value of the sensor is Δε2=ε0-ε2. In order to avoid the sensor buckling and the strain signal changing drastically, Δε2>0 should be ensured at all times.

[0044] Due to the gap between the sensor and the blade, when the blade swings up and down with the same amplitude, the magnitudes of Δε1 and Δε2 are different, and Δε1 > Δε2. This is because: when the blade swings upwards, the sensor relaxes, and the middle section remains separated from the blade. Therefore, the compressive strain of the blade cannot be transmitted to the sensor, and Δε2 is less than the compressive strain on the blade surface. When the blade swings downwards, the gap between the sensor and the blade decreases and they come into contact. Due to the geometric constraints of the blade, the sensor is forced to stretch further. When the initial gap h is sufficiently small, Δε1 is approximately equal to the tensile strain on the blade surface. Furthermore, the tensile and compressive strain values ​​are the same at the same position on symmetrically swinging blades, therefore Δε1 > Δε2.

[0045] Assuming a sinusoidal displacement excitation perpendicular to the blade's spanwise direction is applied to the blade tip, causing the blade to oscillate symmetrically up and down, based on the aforementioned sensor variation characteristics, the sensor output signal is a sinusoidal signal with approximately the same positive half-cycle frequency as the blade surface, and a "peak-clipping" sinusoidal signal during the negative half-cycle, such as... Figure 4 As shown by the solid line.

[0046] When icing occurs, the accumulated ice in the gap causes the sensor to adhere to the blade. At this point, the sensor is tightly pressed against the blade, and the strain detected by the FBG sensor during its up-and-down movement is the actual strain on the blade surface. In this state, Δε1 = Δε2, and the sensor output signal is a standard sine wave, such as... Figure 4 As shown by the dashed line.

[0047] Therefore, by comparing the non-icing signal curve under the current operating conditions to determine whether the current curve has "peak clipping" or disappearance, it is possible to determine whether the wind turbine blades are covered with ice.

[0048] Although embodiments of the invention have been shown and described, it will be understood by those skilled in the art that various changes, modifications, substitutions and alterations can be made to these embodiments without departing from the principles and spirit of the invention, the scope of which is defined by the appended claims and their equivalents.

Claims

1. A method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor, characterized in that, Includes the following steps: Step 1: After pre-tensioning the substrate-type FBG strain sensor, it is pasted onto the leading edge of the wind turbine blade to obtain a bowstring structure composed of the sensor and the wind turbine blade; Step 2: Obtain the strain signal curves of the substrate FBG strain sensor for one complete cycle of wind turbine blade operation under different working conditions without icing using a fiber optic grating demodulator, and mark the features accordingly. Step 3: Obtain the strain signal curve of the substrate FBG strain sensor for one cycle of the wind turbine blade using a fiber optic grating demodulator. Step 4: Compare the strain signal curve of the substrate FBG strain sensor when it is not covered with ice to obtain the "peak clipping" of the current strain signal curve trough and determine the current icing situation. The substrate-type FBG strain sensor in the first step consists of a "π"-shaped substrate and an optical fiber encapsulated in the substrate. The substrate is a glass fiber sheet with a thin middle and thick ends. When the substrate is pasted onto the blade surface, the thicker ends are the pasting sections, which are fixed to the blade by adhesive. The method for obtaining the strain signal curve of a substrate-type FBG strain sensor is as follows: S1: When the blade is not bent or deformed, the sensor is in a tensile state. At this time, the strain value output by the substrate-type FBG strain sensor is recorded as follows: ; S2: As the blade swings downwards, bending and tensile deformations occur on the upper surface of the blade. The sensor is further tightened, increasing the strain. The gap between the sensor and the blade decreases until the sensor contacts the blade. At this point, the strain value of the substrate-type FBG strain sensor is recorded as follows: Therefore, relative to the unbent blade, the strain change value of the substrate FBG strain sensor is... ; S3: The blade swings upward, causing bending deformation and compressive strain on the upper surface of the blade. The sensor relaxes, the strain decreases, and the gap between the substrate FBG strain sensor and the blade increases. At this time, the strain value of the substrate FBG strain sensor is recorded as follows: Therefore, relative to the unbent blade, the strain change value of the substrate FBG strain sensor is... .

2. The method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor according to claim 1, characterized in that: There is a gap between the thinner middle section of the substrate and the blade. ,gap The size of the ice affects the sensor's ice-detecting sensitivity.

3. The method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor according to claim 1, characterized in that: 。 4. The method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor according to claim 1, characterized in that: 。 5. The method for detecting ice on wind turbine blades based on the bowstring mechanical response of an FBG sensor according to claim 1, characterized in that: When icing occurs, the ice buildup in the gaps causes the sensor to adhere to the blade. At this point, the sensor is tightly pressed against the blade, and the strain detected by the substrate-type FBG strain sensor during its up-and-down movement is the true strain on the blade surface. The sensor output signal is a standard sine wave.

Images