Fan blade high-sensitivity vibration sensing method and device based on FBG temperature compensation

Abstract

The invention discloses a fan blade high-sensitivity vibration sensing method and device based on FBG temperature compensation, and relates to the technical field of wind power generation on-line monitoring, and the method comprises the steps: adjusting the voltage through a controller, and enabling a vibration sensor to work in a vibration detection sensitive interval; transmitting a laser signal output by the laser to the vibration sensor in an asymmetric beam splitting ratio structure by using a circulator, and adjusting a coupling relationship between a laser mode of the laser and an optical fiber mode of the vibration sensor based on a mode matching optimization algorithm; demodulating a laser signal reflected by the vibration sensor based on a modal decomposition combined technology; and transmitting the demodulated reflected laser signal to an interferometer to reconstruct a phase diagram, generating a vibration signal by matching a carrier modulator with the interferometer to eliminate initial phase shift, and converting the interfered laser signal into a voltage signal and collecting the voltage signal. The sensor is high in temperature compensation capability, high in vibration detection sensitivity, resistant to electromagnetic interference and high in insulation performance.

Description

Technical Field

[0001] This invention relates to the field of online monitoring technology for wind power generation, and more specifically, to a highly sensitive vibration sensing method and device for wind turbine blades based on FBG temperature compensation. Background Technology

[0002] As a key component of wind turbine units, the condition of wind turbine blades directly affects the power generation efficiency and safety of the turbine. With the rapid development of the wind power industry, the power capacity of wind turbines is constantly increasing, and the blade size is growing, making the blades face more severe challenges during operation. Due to environmental factors such as storms, freezing, and object collisions, as well as material fatigue and structural aging caused by long-term operation, blades may suffer various forms of damage. If these initial damages are not detected and repaired in time, they may further evolve into serious structural failures, not only reducing power generation efficiency but also potentially causing major safety accidents. For such large rotating mechanical equipment, a failure will cause huge economic losses and safety hazards. Therefore, real-time, high-precision vibration monitoring and condition assessment of wind turbine blades has significant engineering value and economic significance.

[0003] Currently, vibration monitoring has become one of the main methods for monitoring the condition of wind turbine blades. Traditional vibration monitoring technologies mainly use accelerometers or displacement sensors, but these sensors are susceptible to electromagnetic interference and have low reliability in harsh environments. In recent years, fiber optic grating (FBG)-based vibration sensing technology has gradually become a new choice for wind turbine blade vibration monitoring due to its advantages such as small size, light weight, strong resistance to electromagnetic interference, corrosion resistance, and long service life. By installing FBG sensors on wind turbine blades, the vibration state of the blades can be monitored in real time, helping maintenance personnel to quickly detect abnormal vibrations, take timely preventive maintenance measures, avoid mechanical failures caused by blade vibration, and improve the reliability and service life of the equipment. Compared with existing electrical vibration testing technologies, the FBG vibration analysis method has no electrical connection with the entire power system, has no adverse impact on the normal operation of the wind turbine, and can safely and reliably achieve online monitoring.

[0004] However, existing FBG-based vibration sensing technologies still face numerous unresolved issues. First, FBG sensors are extremely sensitive to temperature changes; fluctuations in ambient temperature can cause the center wavelength of the fiber optic grating to shift, interfering with the accurate acquisition of vibration signals and reducing measurement precision. Second, traditional signal demodulation methods such as Fast Fourier Transform (FFT) or wavelet transform are ineffective in processing nonlinear and non-stationary vibration signals, struggling to accurately separate complex vibration signals with dense frequency components or aliasing modes. Third, current technologies suffer from deficiencies in optical system coupling efficiency and signal transmission stability, resulting in low signal-to-noise ratios and affecting the sensitivity and accuracy of vibration detection. Finally, the lack of effective signal processing algorithms for noise suppression and feature extraction of the acquired vibration signals makes it difficult to identify and locate vibration anomalies promptly and accurately. These problems severely restrict the application and widespread adoption of FBG vibration sensing technology in wind turbine blade monitoring. Therefore, there is an urgent need to develop a highly sensitive vibration sensing method based on FBG temperature compensation to improve the accuracy, reliability, and applicability of vibration monitoring.

[0005] No effective solutions have yet been proposed to address the problems in the relevant technologies. Summary of the Invention

[0006] To address the problems in related technologies, this invention proposes a highly sensitive vibration sensing method for wind turbine blades based on FBG temperature compensation. This method has the advantages of strong temperature compensation capability, high vibration detection sensitivity, resistance to electromagnetic interference, and strong insulation performance. It also solves the problems of FBG sensors being susceptible to temperature interference, having low signal demodulation accuracy, insufficient noise suppression capability, and difficulty in identifying vibration anomalies in existing technologies.

[0007] Therefore, the specific technical solution adopted by the present invention is as follows:

[0008] According to one aspect of the present invention, a high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation is provided, the method comprising:

[0009] S1. The voltage is adjusted by the controller to optimize the static operating point of the vibration sensor, so that the vibration sensor works within the sensitive range of vibration detection, and the laser is controlled to continuously output a laser signal of a specific wavelength to ensure that the laser signal matches the static operating point.

[0010] S2. The laser signal output by the laser is transmitted to the vibration sensor in an asymmetric beam splitting structure using a circulator, and the coupling relationship between the laser mode of the laser and the fiber mode of the vibration sensor is adjusted based on the mode matching optimization algorithm so that the coupling loss is less than a preset threshold.

[0011] S3. Based on the modal decomposition and joint technology, demodulate the laser signal reflected by the vibration sensor; the modal decomposition and joint technology includes decomposing the laser signal into different intrinsic modal functions and assigning an independent center frequency to each intrinsic modal function to avoid frequency aliasing;

[0012] S4. The demodulated reflected laser signal is transmitted to the interferometer to reconstruct the phase diagram. A vibration signal is generated by the carrier modulator in conjunction with the interferometer to eliminate the initial phase shift. The interferometric laser signal is converted into a voltage signal and acquired based on the coupler and photoelectric converter.

[0013] Furthermore, the method of transmitting the laser signal output by the laser to the vibration sensor using a circulator with an asymmetric beam splitting ratio structure includes: measuring the output spot of the laser based on a beam analyzer; obtaining the input laser mode distribution by recording the spatial intensity distribution of the output spot; and calculating the coupling efficiency based on the fiber mode field distribution and the input laser mode distribution.

[0014] Furthermore, adjusting the coupling relationship between the laser mode and the fiber mode based on the mode matching optimization algorithm includes: establishing a coupling efficiency maximization model based on the optimized matching layer refractive index structure and the principle of minimizing reflection loss; solving the overlap integral of the laser mode distribution and the fiber mode field distribution based on the coupling efficiency maximization model, and adjusting the matching degree between the input laser mode and the fiber mode of the vibration sensor through the gradient descent algorithm; and verifying the signal transmission stability after coupling based on the matching degree adjustment results.

[0015] Furthermore, decomposing the laser signal into different intrinsic mode functions includes: finding local maxima and minima of the reflected signal based on an empirical mode decomposition algorithm; constructing upper and lower envelopes based on the local maxima and minima; subtracting the upper and lower envelopes from the original signal to obtain the residual signal; repeating the residual signal acquisition steps until the high-frequency components of the reflected signal are extracted, and gradually decomposing the remaining low-frequency components into intrinsic mode functions; wherein, the intrinsic mode function includes instantaneous amplitude and instantaneous phase, where instantaneous amplitude represents the intensity of the reflected signal change in the time domain, and instantaneous phase represents the phase change of the reflected signal.

[0016] Furthermore, assigning an independent center frequency to each intrinsic mode function includes: adjusting the frequency of each intrinsic mode function according to the variational mode decomposition algorithm; based on the adjustment result, concentrating the spectrum of each intrinsic mode function within a specific frequency band; and ensuring that the frequency components of each intrinsic mode function have an independent center frequency within the corresponding specific frequency band to avoid frequency aliasing.

[0017] Furthermore, the demodulated reflected laser signal is transmitted to an interferometer to reconstruct the phase diagram. A vibration signal is generated by a carrier modulator in conjunction with the interferometer to eliminate the initial phase shift. Based on a coupler and a photoelectric converter, the interferometric laser signal is converted into a voltage signal and acquired, including: transmitting the laser signal reflected by the vibration sensor to the interferometer and reconstructing the phase diagram using an iterative phase recovery algorithm; applying a sinusoidal voltage signal to the piezoelectric tube in the interferometer through a carrier modulator to generate a sinusoidal periodic vibration signal, and using an autoregressive modeling algorithm based on a long short-term memory network to dynamically model the input voltage to eliminate signal fading caused by the initial phase shift; transmitting the interferometric laser signal to the photoelectric converter using a coupler to convert the laser signal into a voltage signal, and using an adaptive filter to dynamically correct the output voltage; acquiring the voltage signal output by the photoelectric converter based on a signal acquisition processor, estimating the power spectral density of the signal and noise according to a filtering algorithm, and suppressing noise in the acquired voltage signal.

[0018] Furthermore, the laser signal reflected by the vibration sensor is transmitted to the interferometer, and the phase map is reconstructed using an iterative phase retrieval algorithm. This includes: based on the iterative phase retrieval algorithm, repeatedly matching the amplitude and phase of the reflected signal through the frequency and spatial domains; applying Fourier transform to the interferometric image during the iteration process; and performing phase retrieval by combining the reference intensity distribution to obtain the phase map.

[0019] The dynamic correction of the output voltage using an adaptive filter includes: calculating the difference between the desired output and the product of the converter gain and the input signal based on the minimum mean square error adaptive filter to obtain the error function; adding the learning rate and the product of the error function and the input signal to the current converter gain to update the converter gain at the next moment; and performing real-time correction of the output voltage signal based on the updated converter gain to achieve dynamic optimization of photoelectric conversion.

[0020] According to another aspect of the present invention, a high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation is also provided. This high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation includes: a laser for generating a laser signal of a specific wavelength; a circulator for realizing bidirectional optical signal transmission on a single optical fiber; a vibration sensor for amplifying the vibration signal of the wind turbine blade and transmitting the vibration signal along the optical fiber; wherein the fiber grating in the vibration sensor employs wavelength multiplexing technology, and the center wavelengths of adjacent gratings have a preset interval; an interferometer for obtaining the acoustic wave information of the vibration signal; a carrier modulator for generating a carrier wave and cooperating with the piezoelectric tube in the interferometer to generate a carrier signal; a photoelectric converter for converting the laser signal coupled out by the coupler in the interferometer and the laser signal reflected by the vibration sensor into a voltage signal; and a signal acquisition processor for acquiring and analyzing the voltage signal.

[0021] Furthermore, the vibration sensor includes: a base for fixing the sensor, made of stainless steel; an inertial mass block for displacement under the inertial action of vibration, made of brass; a triangular cantilever beam for fixing the inertial mass block, made of stainless steel; and several fiber gratings disposed on the vibration sensing path, used to deform when subjected to stress applied by the inertial mass block. This deformation causes a change in the refractive index distribution inside the fiber gratings, resulting in a corresponding shift in the center wavelength of the reflected light to reflect the vibration. The fiber gratings are fixed between the base and the inertial mass block using a double-point encapsulation, and the signal is processed through differential operations to eliminate the shift in the center wavelength of the fiber gratings caused by temperature changes.

[0022] Furthermore, the interferometer includes: a coupler for coupling laser light; a signal arm for transmitting vibration signals; a piezoelectric tube for generating sinusoidal periodic vibrations under the action of a carrier modulator to eliminate signal fading caused by the initial phase shift of the interferometer; and a Faraday mirror for rotating the polarized light direction by a preset angle and making it orthogonal and perpendicular to the input polarized light with opposite polarization directions, so as to change the polarization state of the input light.

[0023] The beneficial effects of this invention are as follows:

[0024] (1) This invention provides a high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation. By using a double-point encapsulation method to fix the fiber grating between the base and the inertial mass block, and using a triangular cantilever beam to fix the inertial mass block, when the inertial mass block is displaced under the inertial action of external vibration, it will cause the fiber grating to deform, causing the center wavelength of the fiber grating to change. At the same time, this invention effectively reduces the influence of temperature on FBG through differential operation, avoids the chirping phenomenon, significantly improves the sensitivity and accuracy of vibration detection, and solves the technical problem of traditional FBG sensors being greatly affected by temperature interference.

[0025] (2) This invention employs a combined mode decomposition technique to process vibration signals, overcoming the limitations of traditional methods in handling complex vibration signals. Through an empirical mode decomposition algorithm, this invention can decompose complex vibration signals into multiple intrinsic mode functions; and assign an independent center frequency to each mode function, effectively avoiding frequency aliasing. After the signal is transmitted to the interferometer, an iterative phase recovery algorithm is used to reconstruct the phase diagram, and a carrier modulator, in conjunction with a piezoelectric tube, generates sinusoidal vibrations to eliminate the initial phase shift. This multi-level signal processing method significantly improves the system's ability to analyze nonlinear and non-stationary vibration signals, enabling accurate capture of even minute vibration changes in wind turbine blades.

[0026] (3) The fiber optic grating vibration sensor used in this invention has the characteristics of small size, stable physical performance and high insulation, and can be closely integrated with the wind turbine; the entire vibration sensing method has no electrical connection with the power system, has no adverse effect on the normal operation of the wind turbine, and has extremely strong anti-electromagnetic interference capability and cross sensitivity; this technical characteristic makes this invention particularly suitable for use in the complex electromagnetic environment of wind farms, and can safely and reliably realize all-weather online monitoring of wind turbine blade vibration, providing important technical support for preventive maintenance.

[0027] (4) This invention addresses the practical problem that the operational safety and stability of wind turbine blades, as important components for capturing wind energy, directly affect the efficiency and safety of wind turbine units. It provides a highly sensitive vibration sensing device. This device can monitor the vibration state of the blades in real time, helping maintenance personnel to quickly understand whether there is abnormal vibration during wind turbine operation, and thus take timely measures to prevent vibration-induced failures. Through real-time online monitoring of wind turbine vibration signals, this invention effectively extends the service life of the equipment, reduces the risk of accidents caused by abnormal vibration, and improves the overall operating efficiency and reliability of wind power equipment. Attached Figure Description

[0028] To more clearly illustrate the technical solutions in the embodiments of the present invention or the prior art, the drawings used in the embodiments will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For those skilled in the art, other drawings can be obtained based on these drawings without creative effort.

[0029] Figure 1 This is a flowchart illustrating a highly sensitive vibration sensing method for wind turbine blades based on FBG temperature compensation according to an embodiment of the present invention.

[0030] Figure 2 This is a schematic diagram of the structure of a high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation according to an embodiment of the present invention.

[0031] Figure 3 This is a schematic diagram of the vibration sensor in a high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation, according to an embodiment of the present invention. Detailed Implementation

[0032] To further illustrate the various embodiments, the present invention provides accompanying drawings, which are part of the disclosure of the present invention. These drawings are mainly used to illustrate the embodiments and can be used in conjunction with the relevant descriptions in the specification to explain the operating principles of the embodiments. With reference to these drawings, those skilled in the art should be able to understand other possible implementation methods and the advantages of the present invention. The components in the drawings are not drawn to scale, and similar component symbols are generally used to represent similar components.

[0033] According to an embodiment of the present invention, a method and apparatus for high-sensitivity vibration sensing of wind turbine blades based on FBG temperature compensation are provided.

[0034] The present invention will now be further described in conjunction with the accompanying drawings and specific embodiments, such as... Figure 1 As shown, according to an embodiment of the present invention, a high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation is provided. This high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation includes:

[0035] S1. The voltage is adjusted by the controller to optimize the static operating point of the vibration sensor, so that the vibration sensor works within the sensitive range of vibration detection, and the laser is controlled to continuously output a laser signal of a specific wavelength to ensure that the laser signal matches the static operating point.

[0036] S2. The laser signal output by the laser is transmitted to the vibration sensor in an asymmetric beam splitting structure using a circulator, and the coupling relationship between the laser mode of the laser and the fiber mode of the vibration sensor is adjusted based on the mode matching optimization algorithm so that the coupling loss is less than a preset threshold.

[0037] S3. Based on the modal decomposition and joint technology, demodulate the laser signal reflected by the vibration sensor; the modal decomposition and joint technology includes decomposing the laser signal into different intrinsic modal functions and assigning an independent center frequency to each intrinsic modal function to avoid frequency aliasing;

[0038] S4. The demodulated reflected laser signal is transmitted to the interferometer to reconstruct the phase diagram. A vibration signal is generated by the carrier modulator in conjunction with the interferometer to eliminate the initial phase shift. The interferometric laser signal is converted into a voltage signal and acquired based on the coupler and photoelectric converter.

[0039] Specifically, the present invention proposes a highly sensitive vibration sensing method for wind turbine blades based on FBG temperature compensation, the method comprising the following steps;

[0040] Step 1: The controller optimizes the static operating point of the sensor by precisely adjusting the voltage, ensuring that the sensor operates within the sensitive range of vibration detection, and controls the laser to output a stable laser signal, ensuring that the laser signal perfectly matches the operating point of the sensor and avoiding signal drift.

[0041] Step 2: The circulator efficiently transmits the laser signal output by the laser to the fiber Bragg grating (FBG) vibration sensor through an asymmetric beam splitting structure.

[0042] Step 3: After the vibration sensor reflects the laser signal, the signal is demodulated using a combination of Empirical Mode Decomposition (EMD) and Variational Mode Decomposition (VMD) algorithms, replacing traditional FFT or wavelet algorithms. EMD decomposes the reflected signal into several intrinsic mode functions (IMFs), while VMD is used to suppress mode aliasing and improve frequency extraction accuracy.

[0043] Step 4: The laser signal reflected by the vibration sensor is transmitted to the Michelson interferometer, and the iterative phase recovery (Gerchberg-Saxton) algorithm is used to reconstruct the phase map.

[0044] Step 5: The reference arm converts the sinusoidal voltage signal into a vibration signal via a piezoelectric tube using a carrier modulator. To optimize the system response, an autoregressive modeling algorithm based on a Long Short-Term Memory (LSTM) network is used to dynamically model the input voltage, learning the piezoelectric response's variation with temperature and load, thus achieving adaptive adjustment of the piezoelectric drive accuracy.

[0045] Step Six: The coupler transmits the laser signal to the photoelectric converter, which efficiently converts the laser signal into a voltage signal. To improve the accuracy of the signal conversion, a minimum mean square error (LMS) adaptive filter is used to dynamically correct the output voltage.

[0046] Step 7: The signal acquisition processor is used to acquire and process the voltage signal output by the photoelectric converter.

[0047] In one embodiment, transmitting the laser signal output by the laser to the vibration sensor using a circulator with an asymmetric beam splitting ratio structure includes: measuring the output spot of the laser based on a beam analyzer; obtaining the input laser mode distribution by recording the spatial intensity distribution of the output spot; and calculating the coupling efficiency based on the fiber mode field distribution and the input laser mode distribution.

[0048] In one embodiment, adjusting the coupling relationship between the laser mode and the fiber mode based on the mode matching optimization algorithm includes: establishing a coupling efficiency maximization model based on the optimized matching layer refractive index structure and the principle of minimizing reflection loss; solving the overlap integral of the laser mode distribution and the fiber mode field distribution based on the coupling efficiency maximization model, and adjusting the matching degree between the input laser mode and the fiber mode of the vibration sensor through the gradient descent algorithm; and verifying the signal transmission stability after coupling based on the matching degree adjustment result.

[0049] Specifically, in step two, the circulator efficiently transmits the laser signal output from the laser to the fiber Bragg grating (FBG) vibration sensor through an asymmetric beam splitting ratio structure. This coupling mechanism is designed based on an optimized matching layer refractive index structure and the principle of minimizing reflection loss, employing a coupling efficiency maximization model:

[0050] (1)

[0051] In the formula, Indicates the input laser mode distribution. The mode field distribution of the optical fiber is represented by η, where η is the coupling efficiency. By using a mode field matching optimization algorithm, the coupling loss is reduced to less than 0.5 dB, thereby ensuring transmission stability. This is achieved through the following steps:

[0052] Step 2-1: Use a beam analyzer (such as a CCD beam imaging system) to directly measure the output beam of the laser and record its spatial intensity distribution to obtain the input laser mode distribution. Calculate the coupling efficiency using the fiber mode field distribution provided by the manufacturer.

[0053] Step 2-2: Apply the mode field matching optimization algorithm to adjust the matching between the laser mode and the fiber mode, thereby minimizing coupling loss and ensuring that the coupling efficiency is less than 0.5dB;

[0054] Specifically, gradient descent is a type of model-field matching optimization algorithm that iteratively updates parameters to achieve optimal model-field matching.

[0055] Steps 2-3: Verify the stability of the coupled signal transmission to ensure that the system can operate stably over a long period of time and avoid signal loss due to optical mode mismatch.

[0056] In one embodiment, decomposing a laser signal into different intrinsic mode functions includes: finding local maxima and minima of the reflected signal based on an empirical mode decomposition algorithm; constructing upper and lower envelopes based on the local maxima and minima; subtracting the upper and lower envelopes from the original signal to obtain a residual signal; repeating the residual signal acquisition steps until the high-frequency components of the reflected signal are extracted, and gradually decomposing the remaining low-frequency components into intrinsic mode functions; wherein the intrinsic mode functions include instantaneous amplitude and instantaneous phase, the instantaneous amplitude representing the intensity of the reflected signal change in the time domain, and the instantaneous phase representing the phase change of the reflected signal.

[0057] In one embodiment, assigning an independent center frequency to each intrinsic mode function includes: adjusting the frequency of each intrinsic mode function according to a variational mode decomposition algorithm; based on the adjustment result, concentrating the spectrum of each intrinsic mode function within a specific frequency band; and ensuring that the frequency components of each intrinsic mode function have an independent center frequency within the corresponding specific frequency band to avoid frequency aliasing.

[0058] Specifically, in step three, after the vibration sensor reflects the laser signal, the signal is demodulated using an Empirical Mode Decomposition (EMD) combined with Variational Mode Decomposition (VMD) algorithm, replacing the traditional FFT or wavelet algorithm. In this invention, EMD decomposes the reflected signal into several intrinsic mode functions (IMFs), and VMD is used to suppress mode aliasing and improve frequency extraction accuracy. The feature extraction formula is:

[0059] (2)

[0060] In the formula, u k (t) is the envelope of the k-th mode, ω k For the center frequency, The initial phase is defined; specifically, the following steps are included:

[0061] Step 3-1: Use the EMD algorithm to find the local maxima and minima of the signal, construct an envelope using these extrema, and further subtract the envelope from the original signal to obtain the residual signal. Repeat this step until the high-frequency components of the signal are extracted, and the remaining low-frequency components are gradually decomposed into intrinsic mode functions (IMFs), where the intrinsic mode function is described as IMF = A(t)sin( (t)); where A(t) is the instantaneous amplitude, representing the intensity of the signal change in the time domain; (t) represents the instantaneous phase, indicating the phase change of the signal;

[0062] Step 3-2: The VMD algorithm assigns a different center frequency to each frequency band. Specifically, when decomposing the signal, the algorithm adjusts the frequency of each intrinsic mode function so that the spectrum of each mode is concentrated within a specific frequency band. The goal of the algorithm optimization is to minimize the differences between mode frequencies, thereby ensuring that the frequency components of each mode remain independent within their corresponding frequency band and do not overlap with other modes. This optimization process ensures that each mode has an independent center frequency, effectively avoiding frequency aliasing and improving the frequency separation capability of the signal.

[0063] In one embodiment, the demodulated reflected laser signal is transmitted to an interferometer to reconstruct the phase diagram. A vibration signal is generated by a carrier modulator in conjunction with the interferometer to eliminate the initial phase shift. The process involves converting the interferometric laser signal into a voltage signal based on a coupler and a photoelectric converter, and then acquiring the signal. This includes: transmitting the laser signal reflected by the vibration sensor to the interferometer and reconstructing the phase diagram using an iterative phase recovery algorithm; applying a sinusoidal voltage signal to the piezoelectric tube in the interferometer through a carrier modulator to generate a sinusoidal periodic vibration signal, and dynamically modeling the input voltage using an autoregressive modeling algorithm based on a long short-term memory network to eliminate signal fading caused by the initial phase shift; transmitting the interferometric laser signal to the photoelectric converter using a coupler to convert the laser signal into a voltage signal, and dynamically correcting the output voltage using an adaptive filter; and acquiring the voltage signal output by the photoelectric converter based on a signal acquisition processor, estimating the power spectral density of the signal and noise according to a filtering algorithm, and performing noise suppression on the acquired voltage signal.

[0064] In one embodiment, transmitting the laser signal reflected by the vibration sensor to an interferometer and reconstructing the phase map using an iterative phase retrieval algorithm includes: based on the iterative phase retrieval algorithm, repeatedly matching the amplitude and phase of the reflected signal in the frequency and spatial domains; applying Fourier transform to the interferometric image during the iteration process; and performing phase retrieval by combining a reference intensity distribution to obtain the phase map.

[0065] The dynamic correction of the output voltage using an adaptive filter includes: calculating the difference between the desired output and the product of the converter gain and the input signal based on the minimum mean square error adaptive filter to obtain the error function; adding the learning rate and the product of the error function and the input signal to the current converter gain to update the converter gain at the next moment; and performing real-time correction of the output voltage signal based on the updated converter gain to achieve dynamic optimization of photoelectric conversion.

[0066] Specifically, in step four, the laser signal reflected by the vibration sensor is transmitted to a Michelson interferometer, and the Gerchberg–Saxton iterative phase retrieval algorithm is used to reconstruct the phase map. This algorithm improves the clarity and signal-to-noise ratio (SNR) of the interferogram by iteratively matching the amplitude and phase of the reflected signal in the frequency and spatial domains. The expression is as follows:

[0067] (3)

[0068] In the formula, I represents the Fourier transform. (n) For the interferometric image of the nth iteration, A ref This is a reference intensity distribution.

[0069] Specifically, in step five, the reference arm converts the sinusoidal voltage signal into a vibration signal via a piezoelectric tube using a carrier modulator. To optimize the system response, an autoregressive modeling algorithm based on a Long Short-Term Memory (LSTM) network is used to dynamically model the input voltage, learning the variation of the piezoelectric response with temperature and load, thus achieving adaptive adjustment of the piezoelectric drive accuracy. The training objective function is to minimize the mean square error (MSE) between the predicted and actual responses.

[0070] (4)

[0071] In the formula, To predict the output, y i The actual response is represented by n, which is the number of training samples.

[0072] Specifically, in step six, the coupler transmits the laser signal to the photoelectric converter, which efficiently converts the laser signal into a voltage signal. To improve the accuracy of the signal conversion, a minimum mean square error (LMS) adaptive filter is used to dynamically correct the output voltage. The correction algorithm automatically adjusts the converter gain G according to changes in the input optical power to minimize the error function.

[0073] (5)

[0074] In the formula, x(n) is the input signal, d(n) is the desired output, and μ is the learning rate.

[0075] Specifically, in step seven, the signal acquisition processor is used to acquire and process the voltage signal output by the photoelectric converter. To improve the accuracy and reliability of signal processing, the Wiener filtering algorithm is introduced during the processing to smooth and suppress noise in the acquired voltage signal. This algorithm is based on the minimum mean square error criterion and achieves effective suppression of noise interference by estimating the power spectral density of the signal and noise, thereby enhancing the clarity and stability of the signal.

[0076] like Figure 2 and Figure 3 As shown, according to another embodiment of the present invention, a high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation is also provided. This high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation includes:

[0077] A laser is used to generate laser signals of a specific wavelength.

[0078] A circulator is used to enable bidirectional optical signal transmission over a single optical fiber.

[0079] A vibration sensor is used to amplify the vibration signal of the wind turbine blades and transmit the vibration signal along the optical fiber; the fiber grating in the vibration sensor adopts wavelength multiplexing technology, and the center wavelengths of adjacent gratings have a preset interval; an interferometer is used to obtain the acoustic information of the vibration signal.

[0080] A carrier modulator is used to generate a carrier wave and works in conjunction with a piezoelectric tube in an interferometer to generate a carrier signal.

[0081] A photoelectric converter is used to convert the laser signal coupled out by the coupler in the interferometer and the laser signal reflected by the vibration sensor into a voltage signal.

[0082] A signal acquisition processor is used to acquire voltage signals and perform analysis and processing.

[0083] Specifically, such as Figure 2 As shown, the present invention proposes a high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation, comprising a laser, a circulator, a vibration sensor (in this embodiment, a fiber Bragg grating vibration sensor is used), an interferometer (in this embodiment, a Michelson interferometer is used), a carrier modulator, a photoelectric converter, and a signal collection processor.

[0084] The system comprises: a vibration sensor based on a high-sensitivity FBG vibration sensing device with a cantilever beam structure; a laser for generating laser light; a circulator for enabling bidirectional optical signal transmission over a single optical fiber; a sensor for amplifying the vibration signal of the wind turbine blades and transmitting it along the optical fiber, which is wavelength multiplexed with a wavelength interval of 2.4 nm; an interferometer for obtaining the acoustic information of the vibration signal; a carrier modulator for generating a carrier signal that interacts with the piezoelectric tube in the Michelson interferometer; a photoelectric converter for converting the laser signal coupled from the coupler and the laser signal reflected from the vibration sensor into a voltage signal, the difference between which indicates the sensor's static operating point offset; and a signal acquisition processor for acquiring the voltage signal.

[0085] In one embodiment, the vibration sensor includes: a base for fixing the sensor, made of stainless steel; an inertial mass block for displacement under the inertial action of vibration, made of brass; a triangular cantilever beam for fixing the inertial mass block, made of stainless steel; and several fiber gratings disposed on the vibration sensing path for deformation when subjected to stress applied by the inertial mass block. This deformation causes a change in the refractive index distribution inside the fiber gratings, resulting in a corresponding shift in the center wavelength of the reflected light to reflect the vibration. The fiber gratings are fixed between the base and the inertial mass block using a double-point encapsulation, and the signal is processed through differential operations to eliminate the shift in the center wavelength of the fiber gratings caused by temperature changes.

[0086] Specifically, such as Figure 3As shown, the vibration sensor comprises a base, an inertial mass block, a triangular cantilever beam, and two fiber optic gratings (FBG1 and FBG2). The base, used to fix the sensor, is made of stainless steel; the inertial mass block, displaced by the inertia of vibration, is made of brass; the triangular cantilever beam, used to fix the inertial mass block, is made of stainless steel; the fiber optic gratings, deformed by the inertial mass block, change their refractive index, thus changing the wavelength to reflect the vibration, with center wavelengths of 1547.969 nm and 1550.016 nm, respectively.

[0087] Specifically, the resonant frequency of the vibration sensor in this invention is calculated by the following formula:

[0088] (6)

[0089] In the formula, f represents the resonant frequency; K represents the equivalent stiffness of the V-shaped hinge; K f M represents the elastic modulus of the optical fiber; M represents the mass of the mass block.

[0090] Specifically, an FBG (fiber Bragg grating) is disposed in the middle of the optical fiber, and the FBG is located between two inertial mass blocks. Under the action of the inertial mass blocks, the fiber grating deforms, thereby changing its refractive index. The change in refractive index, in turn, changes the center wavelength of the FBG. The vibration sensing device provided by this invention reflects the vibration by changing the wavelength. In the above embodiment, the center wavelengths before and after vibration are 1547.969 nm and 1550.016 nm, respectively.

[0091] Specifically, in this invention, the fiber Bragg grating (FBG) is fixed in the middle of the base and the inertial mass block using a dual-point encapsulation. Differential operations are used to reduce the influence of temperature on the FBG, thereby improving sensitivity. By placing FBGs at both ends of the base and the inertial mass block, and processing the FBG signals at both ends through differential operations, the shift in the center wavelength of the FBG caused by temperature changes can be effectively eliminated. Differential operations calculate the difference in wavelength change between the two FBGs, thereby removing the influence of temperature changes on the FBG wavelength, making the vibration signal detection of the system more accurate.

[0092] In one embodiment, the interferometer includes: a coupler for coupling a laser; a signal arm for transmitting a vibration signal; a piezoelectric tube for generating sinusoidal periodic vibrations under the action of a carrier modulator to eliminate signal fading caused by the initial phase shift of the interferometer; and a Faraday mirror for rotating the polarized light direction by a preset angle and making it orthogonal and perpendicular to the input polarized light with opposite polarization directions to change the polarization state of the input light.

[0093] Specifically, in this embodiment, the interferometer is a Michelson interferometer, including a coupler, a signal arm, a piezoelectric tube, and a Faraday mirror. The coupler is used to couple the laser; the signal arm is used to transmit the vibration signal; the piezoelectric tube generates sinusoidal periodic vibrations under the action of the carrier modulator to eliminate signal fading caused by the initial phase shift of the Michelson interferometer, and a 5kHz high-precision sinusoidal voltage signal is applied to the piezoelectric tube; the Faraday mirror is used to change the polarization state of the input light, so that the direction of the polarized light after passing through the Faraday rotator and the mirror is rotated by 90 degrees, becoming orthogonal and perpendicular to the input polarized light, with opposite polarization directions.

[0094] To facilitate understanding of the above-mentioned technical solution of the present invention, the following is a detailed description using a megawatt-class wind turbine blade vibration monitoring system as an example:

[0095] This embodiment is applied to a 3.6MW offshore wind turbine with blades reaching 68 meters in length, which is prone to severe vibrations in high wind speeds and complex marine environments. To address the vibration monitoring needs of such large wind turbine blades, in step one, the controller first outputs the initial static operating point voltage of the sensor, precisely adjusting it to 2.5V to ensure the system operates within its optimal sensitivity range. Simultaneously, it controls the laser to output a stable laser signal with a wavelength of 1550nm and a power of 10mW. This laser employs a distributed feedback structure, with a linewidth controlled within 0.1nm and temperature stability better than ±0.01℃, providing a high-quality light source foundation for the entire system.

[0096] Then, in step two, the circulator transmits the laser output from the laser to the fiber Bragg grating vibration sensor mounted on the wind turbine blade with an asymmetric beam splitting ratio of 98:2. During transmission, the system uses a mode-matching optimization algorithm to dynamically adjust the coupling relationship between the laser mode and the fiber mode of the vibration sensor, keeping the coupling loss below 0.5dB and effectively improving signal transmission efficiency.

[0097] Next, in step three, when the wind turbine blades vibrate due to external disturbances, the inertial mass block fixed by the triangular cantilever beam in the sensor displaces under the inertial force of the vibration. This causes stress changes in the fiber grating, which is fixed between the base and the inertial mass block using a dual-point encapsulation, resulting in a change in the refractive index distribution inside the grating. Consequently, the center wavelength of the reflected light signal shifts accordingly. The vibration sensor reflects these modulated laser signals, and the system then demodulates these reflected signals. The demodulation process employs a combined mode decomposition technique. This technique first decomposes the laser signal into different intrinsic mode functions based on an empirical mode decomposition algorithm. These functions contain instantaneous amplitude and instantaneous phase information, representing the intensity and phase changes of the reflected signal in the time domain, respectively. Subsequently, the system assigns an independent center frequency to each intrinsic mode function based on a variational mode decomposition algorithm, concentrating its spectrum within a specific frequency band. This effectively avoids frequency aliasing and improves the accuracy of vibration feature identification.

[0098] Secondly, in step four, the demodulated reflected laser signal is transmitted to the Michelson interferometer via a circulator, then passes through the signal arm to the Faraday mirror, and is finally reflected to the coupler. During this process, the wavelength shift of the vibration sensor is converted into a phase shift of the Michelson interferometer, which is calculated using the following formula: The resonant frequency of the vibration sensor is calculated using the following formula:

[0099] (7)

[0100] In the formula, λ represents the phase shift of the Michelson interferometer; n represents the effective refractive index (typically 1.468 in this embodiment); L represents the optical path difference (set to 15 mm); λ represents the center wavelength (1550 nm); Δλ represents the wavelength change.

[0101] Meanwhile, in step five, the reference arm of the Michelson interferometer converts a sinusoidal voltage signal with an amplitude of 3V and a frequency of 1kHz into a vibration signal via a piezoelectric tube, based on a carrier modulator, and then reflects it to the coupler via a Faraday mirror. The Faraday mirror rotates the polarization direction of the light by 45 degrees, effectively changing the polarization state of the input light and improving the system's anti-interference capability.

[0102] Next, in step six, the coupler couples the interfered laser signal to the photoelectric converter, which uses a high-sensitivity PIN photodiode to convert the optical signal into a voltage signal, achieving a conversion efficiency of 0.85 A / W. The system employs an adaptive filtering algorithm based on the minimum mean square error criterion to smooth the output voltage signal and suppress noise, improving the signal-to-noise ratio by approximately 12 dB.

[0103] Finally, in step seven, the signal acquisition processor collects voltage signals at a sampling rate of 10 kHz, reconstructs the phase map through an iterative phase recovery algorithm, and performs noise suppression and feature extraction on the acquired voltage signals, ultimately achieving high-precision monitoring and early warning of abnormalities in the vibration state of wind turbine blades.

[0104] Experimental verification shows that after applying this invention, the system can detect weak vibrations with a sensitivity of up to 10 pm / g, and the frequency response range is 0-1000 Hz. The measurement error is controlled within ±0.5% in the working temperature range of -20℃ to 80℃. This proves that the invention fully meets the vibration sensing and monitoring needs of wind turbine blades in complex environments and can effectively solve the technical problem of traditional FBG sensors being greatly affected by temperature interference.

[0105] The above are merely preferred embodiments of the present invention and are not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of the present invention should be included within the protection scope of the present invention.

Claims

1. A highly sensitive vibration sensing method for wind turbine blades based on FBG temperature compensation, characterized in that, The method includes: S1. The voltage is adjusted by the controller to optimize the static operating point of the vibration sensor, so that the vibration sensor works within the sensitive range of vibration detection, and the laser is controlled to continuously output a laser signal of a specific wavelength to ensure that the laser signal matches the static operating point. S2. The laser signal output by the laser is transmitted to the vibration sensor in an asymmetric beam splitting structure using a circulator, and the coupling relationship between the laser mode of the laser and the fiber mode of the vibration sensor is adjusted based on the mode matching optimization algorithm so that the coupling loss is less than a preset threshold. S3. Based on the modal decomposition and joint technology, demodulate the laser signal reflected by the vibration sensor; the modal decomposition and joint technology includes decomposing the laser signal into different intrinsic modal functions and assigning an independent center frequency to each intrinsic modal function to avoid frequency aliasing; S4. The demodulated reflected laser signal is transmitted to the interferometer to reconstruct the phase diagram. A vibration signal is generated by the carrier modulator in conjunction with the interferometer to eliminate the initial phase shift. The interferometric laser signal is converted into a voltage signal and acquired based on the coupler and photoelectric converter.

2. The high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation according to claim 1, characterized in that, The method of transmitting the laser signal output from the laser to the vibration sensor using a circulator with an asymmetric beam splitting ratio structure includes: The output beam spot of the laser is measured using a beam analyzer. The input laser mode distribution is obtained by recording the spatial intensity distribution of the output light spot; The coupling efficiency is calculated based on the fiber mode field distribution and the input laser mode distribution.

3. The high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation according to claim 2, characterized in that, The adjustment of the coupling relationship between the laser mode and the fiber mode based on the pattern matching optimization algorithm includes: Based on the optimized matching layer refractive index structure and the principle of minimizing reflection loss, a model for maximizing coupling efficiency is established. Based on the coupling efficiency maximization model, the overlap integral of the laser mode distribution and the fiber mode field distribution is solved, and the matching degree between the input laser mode and the fiber mode of the vibration sensor is adjusted by the gradient descent algorithm. Based on the matching degree adjustment results, the stability of the coupled signal transmission is verified.

4. The high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation according to claim 1, characterized in that, The process of decomposing the laser signal into different intrinsic mode functions includes: Based on the empirical mode decomposition algorithm, local maxima and minima of the reflected signal are found; Construct upper and lower envelopes based on local maxima and minima; Subtract the upper and lower envelopes from the original signal to obtain the residual signal; Repeat the residual signal acquisition steps until the high-frequency components of the reflected signal are extracted, and gradually decompose the remaining low-frequency components into intrinsic mode functions; The intrinsic mode function includes instantaneous amplitude and instantaneous phase. The instantaneous amplitude represents the intensity of the reflected signal change in the time domain, and the instantaneous phase represents the phase change of the reflected signal.

5. The high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation according to claim 4, characterized in that, The assignment of an independent center frequency to each intrinsic mode function includes: The frequency of each intrinsic mode function is adjusted according to the variational mode decomposition algorithm. Based on the adjustment results, the spectrum of each intrinsic mode function is concentrated within a specific frequency band; To ensure that the frequency components of each intrinsic mode function have an independent center frequency within their corresponding specific frequency band, frequency aliasing is avoided.

6. The high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation according to claim 1, characterized in that, The process involves transmitting the demodulated reflected laser signal to an interferometer to reconstruct the phase diagram, generating a vibration signal through a carrier modulator in conjunction with the interferometer to eliminate the initial phase shift, and converting the interferometric laser signal into a voltage signal based on a coupler and a photoelectric converter, and then acquiring the following: The laser signal reflected by the vibration sensor is transmitted to the interferometer, and the phase diagram is reconstructed using an iterative phase recovery algorithm; A sinusoidal voltage signal is applied to the piezoelectric tube in the interferometer by a carrier modulator to generate a sinusoidal periodic vibration signal. An autoregressive modeling algorithm based on a long short-term memory network is used to dynamically model the input voltage in order to eliminate signal fading caused by the initial phase shift. The interferometric laser signal is transmitted to a photoelectric converter using a coupler, which converts the laser signal into a voltage signal, and an adaptive filter is used to dynamically correct the output voltage. The voltage signal output from the photoelectric converter is acquired by a signal acquisition processor, and the power spectral density of the signal and noise is estimated according to a filtering algorithm to suppress noise in the acquired voltage signal.

7. The high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation according to claim 6, characterized in that, The process of transmitting the laser signal reflected by the vibration sensor to the interferometer and reconstructing the phase map using an iterative phase retrieval algorithm includes: Based on the iterative phase recovery algorithm, the amplitude and phase of the reflected signal are repeatedly matched in the frequency domain and spatial domain. Applying Fourier transform to the interferometric image during the iterative process; Phase recovery is performed by combining the reference intensity distribution to obtain the phase map; The method of using an adaptive filter to dynamically correct the output voltage includes: Based on the minimum mean square error adaptive filter, the difference between the desired output and the product of the converter gain and the input signal is calculated to obtain the error function; The learning rate is multiplied by the error function and the input signal and added to the current converter gain to update the converter gain at the next time step. The output voltage signal is corrected in real time based on the updated converter gain to achieve dynamic optimization of photoelectric conversion.

8. A high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation, used to implement the high-sensitivity vibration sensing method for wind turbine blades based on FBG temperature compensation as described in any one of claims 1-7, characterized in that, The device includes: A laser is used to generate laser signals of a specific wavelength. A circulator is used to enable bidirectional optical signal transmission over a single optical fiber. A vibration sensor is used to amplify the vibration signal of the wind turbine blades and transmit the vibration signal along the optical fiber; the fiber grating in the vibration sensor adopts wavelength multiplexing technology, and the center wavelengths of adjacent gratings have a preset interval; An interferometer is used to obtain acoustic wave information of vibration signals; A carrier modulator is used to generate a carrier wave and works in conjunction with a piezoelectric tube in an interferometer to generate a carrier signal. A photoelectric converter is used to convert the laser signal coupled out by the coupler in the interferometer and the laser signal reflected by the vibration sensor into a voltage signal. A signal acquisition processor is used to acquire voltage signals and perform analysis and processing.

9. A high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation according to claim 8, characterized in that, The vibration sensor includes: The base, used to fix the sensor, is made of stainless steel; An inertial mass block, used to displace under the inertial action of vibration, is made of brass; A triangular cantilever beam, used to fix an inertial mass block, is made of stainless steel; Several fiber gratings are arranged on the vibration sensing path to generate deformation when subjected to stress by an inertial mass block. This deformation causes a change in the refractive index distribution inside the fiber grating, resulting in a corresponding shift in the center wavelength of the reflected light, thus reflecting the vibration situation. The fiber grating is fixed between the base and the inertial mass block using a two-point encapsulation, and the signal is processed by differential operation to eliminate the offset of the fiber grating center wavelength caused by temperature changes.

10. A high-sensitivity vibration sensing device for wind turbine blades based on FBG temperature compensation according to claim 8, characterized in that, The interferometer includes: Coupler, used for coupling lasers; Signal arm, used to transmit vibration signals; Piezoelectric tubes are used to generate sinusoidal periodic oscillations under the action of a carrier modulator to eliminate signal fading caused by the initial phase shift of the interferometer; A Faraday mirror is used to rotate the direction of polarized light by a preset angle, making it orthogonal and perpendicular to the input polarized light but with the opposite polarization direction, thereby changing the polarization state of the input light.

Images