Fan blade acoustic emission crack detection and evaluation method and system based on morphological filter

Abstract

The invention discloses a morphological filter-based fan blade acoustic emission crack detection and evaluation method and system. The method comprises the following steps: acquiring a blade acoustic emission signal through an acoustic emission sensor, and carrying out low-pass filtering and direct current offset removal pretreatment; dynamically determining the length of morphological structure elements according to the signal variance, and adaptively selecting corrosion or expansion operation to enhance crack characteristics; extracting multi-dimensional crack characteristics including peak values, variances, duration, spectral characteristics and the like; and calculating a crack severity index based on weighted fusion, and evaluating the severity and development situation of the crack in combination with a time sequence change trend. The system comprises a signal acquisition module, an adaptive filtering module, a feature extraction module and a comprehensive evaluation module. According to the invention, high-sensitivity and strong-interference-resistance crack detection and health state monitoring can be realized, and the method is suitable for early damage diagnosis and intelligent operation and maintenance of the fan blade.

Description

Technical Field

[0001] This invention relates to the field of wind power generation technology, specifically to a method and system for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters. Background Technology

[0002] As the core power capture component of a wind turbine, the structural health of wind turbine blades directly determines the power generation efficiency, operational safety, and total life cycle cost of the unit. Blades, typically made of composite materials, are characterized by their large size, complex structure, and high flexibility. They are exposed to variable and even harsh natural environments for extended periods, enduring the combined effects of extreme wind loads, alternating aerodynamic forces, rain erosion, lightning strikes, temperature variations, and material aging. This complex service condition easily leads to initial damage within the blade, such as fiber breakage, matrix cracking, and skin delamination, which gradually evolve into macroscopic cracks. The propagation of these cracks not only disrupts the aerodynamic shape of the blade and reduces power generation efficiency but can also trigger catastrophic fracture accidents, causing enormous economic losses and safety risks. Therefore, early, efficient, and accurate crack detection and health monitoring of wind turbine blades are crucial for ensuring the safe and stable operation of wind farms and reducing maintenance costs.

[0003] Currently, various technical methods have been developed by industry and academia for crack detection in wind turbine blades, mainly including: video and image-based detection techniques. This method typically uses drones or high-powered telescopes equipped with high-definition cameras to photograph or record the blade surface, and then identifies surface cracks through manual visual inspection or image processing algorithms (such as edge detection and image segmentation). Although drone technology improves the accessibility of detection, this method is heavily dependent on lighting and weather conditions, making it difficult to detect internal or micro-cracks. Furthermore, manual interpretation is inefficient and highly subjective, and the processing and analysis of massive amounts of image data also faces significant challenges.

[0004] Vibration analysis-based detection technology collects vibration signals using accelerometers installed at the blade root or main beam, and infers the presence of damage by analyzing changes in structural modal parameters (such as frequency, mode shape, and damping). However, the vibration environment of wind turbine blades during operation is complex, with strong background noise, and the changes in modal parameters caused by cracks are often very weak and easily masked by changes in operating conditions. This results in insufficient sensitivity of the method, high false alarm and false negative rates, and difficulty in achieving reliable diagnosis of early microcracks.

[0005] Traditional non-destructive testing (NDT) techniques include ultrasonic testing, X-ray testing, and thermal imaging. These methods offer high accuracy, but most require shutdown operations and close contact or coupling between the equipment and the blade surface (e.g., ultrasonic testing requires a coupling agent). For wind turbine blades that are tens of meters long, comprehensive testing is time-consuming, labor-intensive, and uneconomical, failing to meet the large-scale, high-frequency inspection needs of modern wind farms.

[0006] Therefore, developing a crack detection and feature extraction method that can overcome the above limitations and has high sensitivity, strong anti-interference ability, high efficiency and good engineering applicability has become an urgent technical need in the field of wind power operation and maintenance. Summary of the Invention

[0007] The purpose of this invention is to provide a method and system for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters, in order to solve problems such as low detection accuracy and reliability, difficulty in balancing detection efficiency and cost, high difficulty in data processing and analysis, and limited adaptability and scalability.

[0008] To achieve the above objectives, the present invention provides a method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters, comprising the following steps: S1. Acoustic emission signals generated by the wind turbine blades during the stress process are captured by the acoustic emission sensor and converted into electrical signals. After the electrical signals are digitized by the analog-to-digital converter, they are subjected to low-pass filtering and DC offset removal processing in sequence to obtain the pre-processed acoustic emission signals. S2. Calculate the standard deviation and variance of the preprocessed acoustic emission signal, dynamically determine the length of the structural element, and adaptively select erosion or dilation operation according to the relationship between the variance and the preset threshold to perform morphological processing on the preprocessed acoustic emission signal. S3. Calculate the crack-related characteristic parameters of the morphologically processed acoustic emission signal, analyze the spectral characteristics of the morphologically processed acoustic emission signal, and extract the crack characteristics of the wind turbine blade. S4. Calculate the crack severity index based on the extracted crack characteristics, and assess the severity and development trend of the crack based on the time series change trend of the crack severity index.

[0009] To optimize the above technical solution, the specific measures also include: In step S1, the specific steps for performing low-pass filtering and DC offset removal are as follows: A low-pass filter is used to remove high-frequency noise; the transfer function is:

[0010] Acoustic emission digital signal Input a low-pass filter to obtain the filtered output. ,calculate Mean:

[0011] in, This is the cutoff frequency of the filter. The filter order; The signal length is used to remove the DC offset by subtracting the average value of the acoustic emission signal, resulting in the preprocessed acoustic emission signal.

[0012] Furthermore, in step S2, the length L of the structural element is determined by the following formula:

[0013] in, The standard deviation of the preprocessed signal; This is the proportionality coefficient; This represents the rounding function; and These are the minimum and maximum allowed lengths of the structuring element, respectively.

[0014] Set a threshold If the variance of the preprocessed acoustic emission signal Then perform morphological erosion operation:

[0015] in, The acoustic emission signal after morphological processing. The erosion operation is represented by z, which represents the preprocessed acoustic emission signal (after DC offset removal); b represents the structuring element in the morphological filter; and m represents the relative time offset index within the sliding window. This method is suitable for eliminating positive spike noise.

[0016] If the variance of the preprocessed acoustic emission signal Then perform a morphological dilation operation:

[0017] Here, ⊕ represents the expansion operation, which is applicable to filling negative gaps or breaks in the signal.

[0018] The extraction of crack features from the wind turbine blades in step S3 includes: Calculate the peak value, mean, and variance of the acoustic emission signal after morphological processing; The spectral characteristics of the acoustic emission signal after morphological processing are analyzed to determine the main frequency components of the crack signal, and the amplitude distribution of the crack at different frequencies is determined based on the main frequency components. The duration of the crack is obtained by calculating the length of time the signal exceeds a preset threshold. The location of the crack is estimated based on the time difference of arrival (TOA) between multiple acoustic emission sensors. According to signal energy Alternatively, cumulative counting can be used to assess crack size.

[0019] The crack severity index in step S4 The calculation formula is as follows: (This is a weighted fusion based on peak value, variance, and crack duration.)

[0020] in, Indicates the peak value of the crack. This represents the variance of the filtered signal. Indicates the duration of the crack; , , This represents the weighting coefficient.

[0021] The method for assessing the severity and development trend of the crack in step S4 is as follows: Record at different monitoring times The calculated crack severity indices are denoted as follows: ; Calculate the rate of change based on two consecutive measurements:

[0022] According to the rate of change The symbol and size of the crack are used to classify the crack development state into rapid propagation, slow propagation, stable or weakened state, and output the corresponding evaluation results.

[0023] As another important technical solution, the present invention also provides a wind turbine blade acoustic emission crack detection and evaluation system based on morphological filters, comprising: The acoustic emission signal acquisition and preprocessing module is used to capture the acoustic emission signal generated by the wind turbine blades during the stress process through the acoustic emission sensor and convert it into an electrical signal. After the electrical signal is digitized by the analog-to-digital converter, it is subjected to low-pass filtering and DC offset removal processing in sequence to obtain the preprocessed acoustic emission signal. The adaptive morphological filtering module is used to calculate the standard deviation and variance of the preprocessed acoustic emission signal, dynamically determine the length of the structuring element, and adaptively select erosion or dilation operation according to the relationship between the variance and the preset threshold to perform morphological processing on the preprocessed acoustic emission signal. The multi-feature extraction module is used to calculate the crack-related feature parameters of the morphologically processed acoustic emission signal, analyze the spectral characteristics of the morphologically processed acoustic emission signal, and extract the crack features of the wind turbine blade. The crack comprehensive assessment and prediction module calculates the crack severity index based on the extracted crack characteristics, and assesses the severity and development trend of the crack based on the time series change trend of the crack severity index.

[0024] The present invention also proposes an electronic device, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein when the processor executes the computer program, it implements the method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters as described above.

[0025] The present invention also proposes a computer-readable storage medium storing a computer program that enables a computer to execute the above-described method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters.

[0026] Compared with the prior art, the beneficial effects of the present invention are: 1. The morphological filter of this invention can effectively filter out noise and interference signals, highlight crack features, and improve the accuracy and reliability of detection.

[0027] 2. The acoustic emission technology in this invention has the ability to monitor in real time. When combined with a morphological filter, it can realize continuous online monitoring of wind turbine blade cracks and detect potential safety hazards such as crack propagation in a timely manner.

[0028] 3. The method and equipment of the present invention are simple and easy to operate, which can reduce detection costs and improve detection efficiency, and are suitable for rapid detection of large-scale wind farms.

[0029] 4. By adjusting the parameters and algorithm of the morphological filter, this invention can adapt to the crack detection needs of wind turbine blades of different types and materials; at the same time, this invention can also be extended to other fields and industries that require crack detection. Attached Figure Description

[0030] Figure 1 The present invention provides a principle block diagram of the acoustic emission crack detection method for wind turbine blades based on morphological filters. Detailed Implementation

[0031] The present invention will be further described in detail below through specific embodiments, but it should not be construed as limiting the scope of the subject matter of the present invention to the following embodiments. All technologies implemented based on the above content of the present invention fall within the scope of the present invention.

[0032] like Figure 1As shown, this invention provides a method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters, comprising the following steps: S1. Acoustic emission signals generated by the wind turbine blades during the stress process are captured by the acoustic emission sensor and converted into electrical signals. After the electrical signals are digitized by the analog-to-digital converter, they are subjected to low-pass filtering and DC offset removal processing in sequence to obtain the pre-processed acoustic emission signals.

[0033] In some implementations, one or more acoustic emission sensors arranged on the surface of the wind turbine blades are used to collect elastic wave signals generated in real time by the deformation of the blades due to stress or micro-fractures of the material during operation.

[0034] Preferably, the sensor is a resonant or broadband acoustic emission sensor with an effective response frequency band covering 50kHz to 500kHz, preferably with a center frequency of 100kHz to 200kHz, to match the high-frequency elastic wave characteristics induced by composite material cracks and suppress low-frequency mechanical vibration interference.

[0035] The acquired analog signal is amplified by a preamplifier and then digitized by an analog-to-digital converter (ADC) with a sampling rate of at least 1 MS / s (millions of times per second) and a resolution of at least 16 bits to obtain a discrete-time signal. .

[0036] The digital signal is then preprocessed: Using a cutoff frequency of A 4th-order Butterworth low-pass filter with a frequency of 300kHz suppresses high-frequency noise. Calculate the filtered signal arithmetic mean And by subtracting this mean, a preprocessed signal with DC offset removed is obtained:

[0037] S2. Calculate the standard deviation and variance of the preprocessed acoustic emission signal, dynamically determine the length of the structuring element, and adaptively select erosion or dilation operation according to the relationship between the variance and the preset threshold to perform morphological processing on the preprocessed acoustic emission signal.

[0038] In some implementations, the length of the structural element Dynamically adjust according to the following formula:

[0039] in, The standard deviation of the preprocessed signal; the scaling factor. The value range is (0.8, 2.0); in this embodiment, k=1.2; This represents the rounding function; =3; =50 is used to prevent filtering failure caused by windows that are too short or too long.

[0040] Structural elements used A one-dimensional flat structuring element is defined as having a length of The zero-value sequence:

[0041] This structuring element is used for local extremum search operations within a sliding window.

[0042] In some implementations, the preprocessed signal is calculated. Standard deviation And set a preset threshold. (For example, take 0.6 times the median of the variance of all historical data) to determine the noise level of the current signal.

[0043] When performing filtering operations, a symmetrical sliding window mechanism is used: If the variance of the preprocessed acoustic emission signal This indicates that the signal contains a lot of sharp pulse noise, and morphological erosion should be selected to suppress positive spikes.

[0044] in, The acoustic emission signal after morphological processing. The erosion operation is represented by z, which represents the preprocessed acoustic emission signal (after DC offset removal); b represents the structuring element in the morphological filter; and m represents the relative time offset index within the sliding window. This method is suitable for eliminating positive spike noise.

[0045] If the variance of the preprocessed acoustic emission signal This indicates that there are local gaps or low-amplitude fluctuations in the signal, and an expansion operation should be selected to fill the signal gaps.

[0046] Here, ⊕ represents the expansion operation, which is applicable to filling negative gaps or breaks in the signal.

[0047] Output morphologically processed acoustic emission signal Its signal-to-noise ratio is significantly improved, and crack-related features are more prominent.

[0048] S3. Calculate the crack-related characteristic parameters of the morphologically processed acoustic emission signal, analyze the spectral characteristics of the morphologically processed acoustic emission signal, and extract the crack characteristics of the wind turbine blade.

[0049] In some implementations, the signal is based on morphological processing. Extract the following crack-related features: Peak value: It reflects the energy intensity of a single event.

[0050] Mean: It characterizes the overall shift trend of the signal and is used to determine whether there is continuous microcrack activity.

[0051] variance: This measures the activity level of signal fluctuations.

[0052] Analyze the spectral characteristics of the acoustic emission signal after morphological processing to determine the main frequency components of the crack signal. Based on the main frequency components Determine the amplitude distribution of the crack at different frequencies.

[0053] Set amplitude threshold like =0.1), detect the start and end times of the event, and calculate the crack duration. .

[0054] If multiple sensors are used, the location of the crack can be estimated using the Time Difference of Arrival (TOA) algorithm.

[0055] Calculate signal energy Alternatively, cumulative counts can be used as an indirect indicator of crack size.

[0056] S4. Calculate the crack severity index based on the extracted crack characteristics, and assess the severity and development trend of the crack based on the time series change trend of the crack severity index.

[0057] In some implementations, the extracted key features are input into the crack severity index model. A weighted index is constructed by selecting peak value, variance, and duration.

[0058] in, Indicates the peak value of the crack. This represents the variance of the filtered signal. Indicates the duration of the crack; the weighting coefficients satisfy... Recommended value =0.4, =0.4, =0.2, which can be calibrated according to different unit types or operating environments.

[0059] The system executes the above process periodically and records data at different monitoring times. The calculated crack severity indices are denoted as follows: .

[0060] Calculate the rate of change between two consecutive evaluation results as an indicator of development trend: according to The symbol and size of the crack are used to classify the crack development state into rapid propagation, slow propagation, stable or weakened state, and output the corresponding evaluation results.

[0061] In some implementations, two preset thresholds are set: Rapidly expand threshold =0.05 / min; Stability determination threshold =0.01 / min; according to Classify states based on their numerical range:

[0062] In some implementations, the evaluation results can be uploaded to the wind farm monitoring system via a local display unit or a remote communication module to achieve intelligent operation and maintenance decision support.

[0063] In another embodiment of the present invention, a wind turbine blade acoustic emission crack detection and evaluation system based on morphological filters is proposed, comprising: The acoustic emission signal acquisition and preprocessing module is used to capture the acoustic emission signal generated by the wind turbine blades during the stress process through the acoustic emission sensor and convert it into an electrical signal. After the electrical signal is digitized by the analog-to-digital converter, it is subjected to low-pass filtering and DC offset removal processing in sequence to obtain the preprocessed acoustic emission signal.

[0064] An adaptive morphological filtering module is used to calculate the variance of the preprocessed acoustic emission signal, dynamically determine the length of the structuring element based on the relationship between the variance value and a preset threshold, and adaptively select erosion or dilation operations to perform morphological processing on the preprocessed acoustic emission signal.

[0065] The multi-feature extraction module is used to calculate the peak value, mean, and variance of the morphologically processed acoustic emission signal as crack-related feature parameters, and analyze the spectral characteristics of the morphologically processed acoustic emission signal to extract the crack features of the wind turbine blade.

[0066] The crack comprehensive assessment and prediction module is used to calculate the crack severity index based on the extracted crack characteristics, and to assess the severity and development trend of the crack based on the time series change trend of the crack severity index.

[0067] In another embodiment of the present invention, an electronic device is proposed, comprising: a memory, a processor, and a computer program stored in the memory and executable on the processor. When the processor executes the computer program, it implements the method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters as described above.

[0068] In another embodiment of the present invention, a computer-readable storage medium is provided, storing a computer program that causes a computer to execute a method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters, as described above.

[0069] In the embodiments disclosed in this application, a computer storage medium may be a tangible medium that may contain or store programs for use by or in conjunction with an instruction execution system, apparatus, or device. The computer storage medium may include, but is not limited to, electronic, magnetic, optical, electromagnetic, infrared, or semiconductor systems, apparatus, or devices, or any suitable combination of the foregoing. More specific examples of computer storage media include electrical connections based on one or more wires, portable computer disks, hard disks, random access memory (RAM), read-only memory (ROM), erasable programmable read-only memory (EPROM or flash memory), optical fibers, portable compact disk read-only memory (CD-ROM), optical storage devices, magnetic storage devices, or any suitable combination of the foregoing.

[0070] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Any simple modifications, equivalent substitutions, and improvements made by those skilled in the art to the above embodiments without departing from the scope of the technical solution of the present invention, based on the technical essence of the present invention, shall still fall within the protection scope of the technical solution of the present invention.

Claims

1. A method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters, characterized in that, Includes the following steps: S1. Acoustic emission signals generated by the wind turbine blades during the stress process are captured by the acoustic emission sensor and converted into electrical signals. After the electrical signals are digitized by the analog-to-digital converter, they are subjected to low-pass filtering and DC offset removal processing in sequence to obtain the pre-processed acoustic emission signals. S2. Calculate the standard deviation and variance of the preprocessed acoustic emission signal, dynamically determine the length of the structural element, and adaptively select erosion or dilation operation according to the relationship between the variance and the preset threshold to perform morphological processing on the preprocessed acoustic emission signal. S3. Calculate the crack-related characteristic parameters of the morphologically processed acoustic emission signal, analyze the spectral characteristics of the morphologically processed acoustic emission signal, and extract the crack characteristics of the wind turbine blade. S4. Calculate the crack severity index based on the extracted crack characteristics, and assess the severity and development trend of the crack based on the time series change trend of the crack severity index.

2. The method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters according to claim 1, characterized in that: In step S1, the specific steps for performing low-pass filtering and DC offset removal are as follows: A low-pass filter is used to remove high-frequency noise; the transfer function is: Acoustic emission digital signal Input a low-pass filter to obtain the filtered output. ,calculate Mean: in, This is the cutoff frequency of the filter. The filter order; The signal length is used to remove the DC offset by subtracting the average value of the acoustic emission signal, resulting in the preprocessed acoustic emission signal. 。 3. The method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters according to claim 1, characterized in that: In step S2, the length L of the structural element is determined by the following formula: in, The standard deviation of the preprocessed signal; This is the proportionality coefficient; This represents the rounding function; and These are the minimum and maximum allowed lengths of the structuring element, respectively.

4. The method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters according to claim 3, characterized in that: Set threshold If the variance of the preprocessed acoustic emission signal Then perform morphological erosion operation: in, This represents the filtered acoustic emission signal. denoted by erosion operation, z represents the preprocessed acoustic emission signal; b represents the structuring element in morphological filtering; m represents the relative time offset index within the sliding window. If the variance of the preprocessed acoustic emission signal Then perform a morphological dilation operation: Here, ⊕ represents the expansion operation.

5. The method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters according to claim 1, characterized in that: The extraction of crack features from the wind turbine blades in step S3 includes: Calculate the peak value, mean, and variance of the acoustic emission signal after morphological processing; The spectral characteristics of the acoustic emission signal after morphological processing are analyzed to determine the main frequency components of the crack signal, and the amplitude distribution of the crack at different frequencies is determined based on the main frequency components. The duration of the crack is obtained by calculating the length of time the signal exceeds a preset threshold. The location of the crack is estimated based on the time difference of arrival (TOA) between multiple acoustic emission sensors. According to signal energy Alternatively, cumulative counting can be used to assess crack size.

6. The method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters according to claim 5, characterized in that: The crack severity index in step S4 The calculation formula is as follows: (This is a weighted fusion based on peak value, variance, and crack duration.) in, Indicates the peak value of the crack. This represents the variance of the filtered signal. Indicates the duration of the crack; , , This represents the weighting coefficient.

7. The method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters according to claim 6, characterized in that: The method for assessing the severity and development trend of the crack in step S4 is as follows: Record at different monitoring times The calculated crack severity indices are denoted as follows: ; Calculate the rate of change based on two consecutive measurements: According to the rate of change The sign and size of the crack are used to classify the crack development state into rapid propagation, slow propagation, stable or weakened state, and the corresponding evaluation results are output.

8. A system for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters, characterized in that, include: The acoustic emission signal acquisition and preprocessing module is used to capture the acoustic emission signal generated by the wind turbine blades during the stress process through the acoustic emission sensor and convert it into an electrical signal. After the electrical signal is digitized by the analog-to-digital converter, it is subjected to low-pass filtering and DC offset removal processing in sequence to obtain the preprocessed acoustic emission signal. The adaptive morphological filtering module is used to calculate the standard deviation and variance of the preprocessed acoustic emission signal, dynamically determine the length of the structuring element, and adaptively select erosion or dilation operation according to the relationship between the variance and the preset threshold to perform morphological processing on the preprocessed acoustic emission signal. The multi-feature extraction module is used to calculate the crack-related feature parameters of the morphologically processed acoustic emission signal, analyze the spectral characteristics of the morphologically processed acoustic emission signal, and extract the crack features of the wind turbine blade. The crack comprehensive assessment and prediction module calculates the crack severity index based on the extracted crack characteristics, and assesses the severity and development trend of the crack based on the time series change trend of the crack severity index.

9. An electronic device, characterized in that, include: The present invention includes a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the computer program to implement a method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters as described in any one of claims 1 to 7.

10. A computer-readable storage medium having a computer program stored thereon, characterized in that: The computer program causes the computer to execute a method for detecting and evaluating acoustic emission cracks in wind turbine blades based on morphological filters, as described in any one of claims 1 to 7.

Images