Anchor rod quality defect detection method

Abstract

The invention relates to the technical field of data processing, in particular to an anchor rod quality defect detection method, which comprises the following steps of: aiming at any to-be-detected anchor rod, performing sound wave detection on any to-be-detected anchor rod to obtain an initial reflection signal; according to the distribution characteristics of the signals in the initial reflection signals and the influence degree of the noise signals, obtaining a decomposition layer number adjustment coefficient when wavelet denoising is carried out on the initial reflection signals; obtaining a decomposition layer number when wavelet denoising is carried out on the initial reflection signal, adjusting the decomposition layer number according to a decomposition layer number adjustment coefficient, and obtaining an optimal decomposition layer number when wavelet denoising is carried out on the initial reflection signal; the optimal number of decomposition layers is used as the number of decomposition layers of wavelet denoising, wavelet denoising is carried out on the initial reflection signal to obtain a denoised signal, and quality defect detection is carried out on any anchor rod to be detected according to the denoised signal, so that the accuracy of anchor rod quality defect detection is improved.

Description

Technical Field

[0001] The invention relates to the technical field of data processing, and in particular to a method for detecting quality defects of anchor rods. Background Art

[0002] Anchor rods are a rod system structure for reinforcing rock and soil. As a tensile member deep in the stratum, one end is connected to the engineering structure and the other end is deep in the stratum. The reinforcement effect of anchor rods not only helps to improve the bearing capacity of the soil, but also effectively resists the scouring and erosion of groundwater, ensuring the safety of the engineering structure. Therefore, the quality defect detection of anchor rods is an important part of ensuring the safety and stability of building structures.

[0003] The traditional method of anchor bolt quality defect detection usually uses the acoustic wave reflection method, which is a commonly used non-destructive testing technology. The vibration source excites the impact elastic wave at the top of the anchor bolt. The elastic wave propagates along the axis of the anchor bolt. When it encounters the interface with wave impedance change, it reflects and forms a reflected wave. The anchor bolt length and anchor defect location are evaluated according to the arrival time and energy attenuation of the reflected wave. However, due to the influence of instrument settings, environmental interference and anchor bolt surface conditions, the reflected wave obtained during acoustic wave detection may be subject to noise interference, resulting in inaccurate judgment of anchor bolt defects. Therefore, it is necessary to denoise the obtained reflected wave.

[0004] Traditional methods often use wavelet denoising to remove noise in acoustic wave detection. First, the appropriate wavelet basis function and decomposition layer number are selected to perform wavelet transform on the initial reflection signal obtained by acoustic wave detection to obtain wavelet coefficients, and then the appropriate threshold function is selected to process the wavelet coefficients to remove the noise signal contained in the initial reflection signal. Finally, the processed wavelet coefficients are subjected to inverse wavelet transform to obtain the denoised signal. However, in the process of wavelet denoising, the setting of the number of decomposition layers has a great influence on the denoising results. A small number of layers may not be able to capture and remove the noise components in the signal, and a large number of layers may cause signal distortion. The setting of the number of decomposition layers of wavelet denoising in traditional methods usually relies on experience to select the number of layers. The frequency of the initial reflection signal obtained by the anchor in acoustic wave detection is generally concentrated in the low-frequency part. At this time, a larger number of decomposition layers is required for more detailed analysis. Therefore, the selection of the number of decomposition layers in wavelet denoising under the traditional method is relatively small compared to the number of decomposition layers required for anchor quality defect detection, which may lead to poor denoising results for the initial reflection signal, making the acoustic wave reflection method unable to detect the anchor quality more accurately.

[0005] Therefore, how to accurately obtain the number of decomposition layers during wavelet denoising so that it can have a better denoising effect in anchor quality defect detection has become an urgent problem to be solved. Summary of the invention

[0006] In view of this, an embodiment of the present invention provides a method for detecting anchor quality defects to solve the problem of how to accurately obtain the number of decomposition layers during wavelet denoising so as to achieve better denoising effect in anchor quality defect detection.

[0007] An embodiment of the present invention provides a method for detecting anchor quality defects, the method comprising the following steps: For any anchor rod to be detected, performing acoustic wave detection on the anchor rod to be detected to obtain an initial reflection signal; According to the distribution characteristics of the signal in the initial reflection signal, the signal complexity of the initial reflection signal is obtained, according to the influence degree of the noise signal in the initial reflection signal, the noise proportion of the initial reflection signal is obtained, and according to the signal complexity and noise proportion of the initial reflection signal, the decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal is obtained; Obtaining the number of decomposition layers when performing wavelet denoising on the initial reflection signal, and adjusting the number of decomposition layers according to the decomposition layer number adjustment coefficient to obtain the optimal number of decomposition layers for performing wavelet denoising on the initial reflection signal; The optimal decomposition layer number is used as the decomposition layer number of wavelet denoising, wavelet denoising is performed on the initial reflection signal to obtain a denoised signal, and quality defect detection is performed on any anchor rod to be inspected based on the denoised signal.

[0008] Preferably, acquiring the signal complexity of the initial reflected signal according to the distribution characteristics of the signal in the initial reflected signal includes: Performing wavelet denoising on the initial reflection signal to obtain a comparison signal, composing the comparison signal into a comparison signal sequence, obtaining sample entropy of the comparison signal sequence, and normalizing the sample entropy to obtain a first complexity of the initial reflection signal; Constructing a frequency spectrum of the initial reflection signal, wherein the abscissa of the frequency spectrum is the frequency and the ordinate is the signal amplitude, obtaining the bandwidth of the initial reflection signal according to the frequency spectrum, and performing normalization processing on the bandwidth to obtain the second complexity of the initial reflection signal; A weighted sum is performed on the first complexity and the second complexity to obtain the signal complexity of the initial reflected signal.

[0009] Preferably, obtaining the noise ratio of the initial reflected signal according to the influence degree of the noise signal in the initial reflected signal includes: Performing Fourier transform on the initial reflected signal to obtain a power spectrum curve graph of the initial reflected signal, wherein the abscissa of the power spectrum curve graph is frequency and the ordinate is power; Obtaining the power spectrum density corresponding to each frequency in the spectrum graph in the power spectrum curve graph, and obtaining the noise energy proportion of the initial reflected signal according to the difference in the power spectrum density corresponding to each frequency in the spectrum graph; According to the power spectrum curve, the average power of the initial reflected signal is obtained, which is recorded as the initial reflected signal average power; the power spectrum curve of the comparison signal is obtained; according to the power spectrum curve of the comparison signal, the average power of the comparison signal is obtained, which is recorded as the comparison signal average power; according to the difference between the initial reflected signal average power and the comparison signal average power, the noise power ratio of the initial reflected signal is obtained; A weighted sum is performed on the noise energy proportion and the noise power proportion to obtain the noise proportion of the initial reflected signal.

[0010] Preferably, obtaining the noise energy proportion of the initial reflected signal according to the difference in power spectrum density corresponding to each frequency in the spectrum diagram includes: Obtaining a maximum power spectrum density from the power spectrum densities corresponding to each frequency in the spectrum graph, and composing a power spectrum density sequence from all power spectrum densities except the maximum power spectrum density; Record the frequency corresponding to the maximum power spectrum density as the target frequency, and for any power spectrum density in the power spectrum density sequence, obtain the absolute value of the frequency difference between the frequency corresponding to any power spectrum density and the target frequency, obtain the proportion of the absolute value of the frequency difference in the absolute values ​​of the frequency differences corresponding to all power spectrum densities in the power spectrum density sequence, and obtain the weight coefficient of any power spectrum density; Obtain a weight coefficient for each power spectral density in the power spectral density sequence, perform weighted summation on all power spectral densities in the power spectral density sequence to obtain a noise energy value, normalize the ratio of the noise energy value to the maximum power spectral density, and obtain the noise energy ratio of the initial reflection signal.

[0011] Preferably, obtaining the noise power ratio of the initial reflected signal according to the difference between the average power of the initial reflected signal and the average power of the comparison signal includes: The average power difference between the average power of the initial reflected signal and the average power of the comparison signal is calculated, and the ratio of the average power difference to the average power of the initial reflected signal is normalized to obtain the noise power ratio of the initial reflected signal.

[0012] Preferably, the step of obtaining a decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal according to the signal complexity and noise proportion of the initial reflection signal includes: A weighted sum is taken for the signal complexity and noise proportion of the initial reflection signal to obtain a decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal.

[0013] Preferably, the step of adjusting the number of decomposition layers according to the decomposition layer number adjustment coefficient to obtain the optimal number of decomposition layers for performing wavelet denoising on the initial reflection signal comprises: When the decomposition layer number adjustment coefficient is greater than a preset decomposition layer number adjustment coefficient threshold, the decomposition layer number adjustment coefficient is used as an independent variable of a Sigmoid function to obtain a Sigmoid function value; The addition result of the constant 1 and the Sigmoid function value is obtained to obtain the expansion coefficient of the decomposition layer number, and the product of the expansion coefficient and the decomposition layer number is rounded up to obtain the optimal decomposition layer number for wavelet denoising of the initial reflection signal.

[0014] Preferably, the step of adjusting the number of decomposition layers according to the decomposition layer number adjustment coefficient to obtain the optimal number of decomposition layers for wavelet denoising of the initial reflection signal further includes: When the decomposition layer number adjustment coefficient is less than or equal to a preset decomposition layer number adjustment coefficient threshold, the decomposition layer number is used as the optimal decomposition layer number for wavelet denoising of the initial reflection signal.

[0015] Compared with the prior art, the embodiments of the present invention have the following beneficial effects: The present invention performs acoustic wave detection on any anchor rod to be detected to obtain an initial reflection signal; obtains the signal complexity of the initial reflection signal according to the distribution characteristics of the signal in the initial reflection signal, obtains the noise proportion of the initial reflection signal according to the influence of the noise signal in the initial reflection signal, and obtains the decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal according to the signal complexity and noise proportion of the initial reflection signal; obtains the decomposition layer number when performing wavelet denoising on the initial reflection signal, adjusts the decomposition layer number according to the decomposition layer number adjustment coefficient, and obtains the optimal decomposition layer number for wavelet denoising on the initial reflection signal; uses the optimal decomposition layer number as the decomposition layer number for wavelet denoising, performs wavelet denoising on the initial reflection signal, obtains a denoised signal, and performs quality defect detection on any anchor rod to be detected according to the denoised signal. Among them, according to the distribution characteristics of the signal in the obtained initial reflection signal and the influence of the noise signal, the decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal is obtained, and then the decomposition layer number is adjusted according to the decomposition layer number adjustment coefficient, and the optimal decomposition layer number for wavelet denoising on the initial reflection signal is accurately obtained, which reduces the poor noise removal effect caused by the traditional method of relying on experience to select the wavelet denoising decomposition layer number, and the optimal decomposition layer number is more dependent on the characteristics of the initial reflection signal itself, so that the acoustic wave reflection method can more accurately detect anchor quality defects, thereby improving the accuracy of anchor quality defect detection. BRIEF DESCRIPTION OF THE DRAWINGS

[0016] In order to more clearly illustrate the technical solutions in the embodiments of the present invention, the drawings required for use in the embodiments or the description of the prior art will be briefly introduced below. Obviously, the drawings described below are only some embodiments of the present invention. For ordinary technicians in this field, other drawings can be obtained based on these drawings without paying creative labor.

[0017] Figure 1 is a flow chart of a method for detecting anchor quality defects provided by the first embodiment of the present invention; Figure 2 This is a physical diagram of an anchor rod provided in the first embodiment of the present invention; Figure 3 This is an example diagram of a spectrum diagram provided in Embodiment 1 of the present invention. DETAILED DESCRIPTION

[0018] Embodiments of the present disclosure are described in detail below, and examples of the embodiments are shown in the accompanying drawings. The embodiments described below with reference to the accompanying drawings are exemplary and are intended to be used to explain the present disclosure, but should not be understood as limiting the present disclosure.

[0019] It should be noted that the terms "first", "second", etc. in the specification of the present disclosure and the above-mentioned drawings are used to distinguish similar objects, and are not necessarily used to describe a specific order or sequence. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present disclosure described herein can be implemented in an order other than those illustrated or described herein. The embodiments described in the following exemplary embodiments do not represent all embodiments consistent with the present disclosure. Instead, they are merely examples of devices and methods consistent with some aspects of the present disclosure.

[0020] In order to illustrate the technical solution of the present invention, specific embodiments are provided below for illustration.

[0021] See also Figure 1 , is a flow chart of a method for detecting anchor quality defects provided by the first embodiment of the present invention, such as Figure 1 As shown, the method may include: Step S101 : for any anchor rod to be detected, perform acoustic wave detection on the anchor rod to be detected to obtain an initial reflection signal.

[0022] Anchor support refers to implanting a rod into the rock mass to modify the original mechanical state of the rock mass, stabilize the broken and discontinuous rock mass, sandstone layer, etc., and make it a whole with high strength and good stability, so as to achieve the purpose of supporting the tunnel. The actual picture of the anchor is as follows Figure 2 shown.

[0023] In order to detect the quality defects of anchor rods, the acoustic wave reflection method is usually used to excite the impact elastic wave at the top of the anchor rod through a vibration source (supermagnetic vibration source or hand hammer). The elastic wave propagates along the axis of the anchor rod. When it encounters an interface with a wave impedance change (grouting defect or rod bottom), it is reflected to form a reflected wave. The anchor rod length and anchor defect position are evaluated according to the arrival time and energy attenuation of the reflected wave. The acoustic wave reflection method belongs to the existing technology and will not be described here. However, due to the influence of instrument settings, environmental interference, and the surface state of the anchor rod, the reflected wave obtained during the acoustic wave detection may be subject to noise interference, resulting in inaccurate judgment of anchor rod defects. Therefore, it is necessary to denoise the obtained reflected wave.

[0024] Traditional methods often use wavelet denoising to remove noise in acoustic wave detection. First, the appropriate wavelet basis function and decomposition layer number are selected to perform wavelet transform on the initial reflection signal obtained by acoustic wave detection to obtain wavelet coefficients, and then the appropriate threshold function is selected to process the wavelet coefficients to remove the noise signal contained in the initial reflection signal. Finally, the processed wavelet coefficients are subjected to inverse wavelet transform to obtain the denoised signal. However, in the process of wavelet denoising, the setting of the number of decomposition layers has a great influence on the denoising results. A small number of layers may not be able to capture and remove the noise components in the signal, and a large number of layers may cause signal distortion. The setting of the number of decomposition layers of wavelet denoising in traditional methods usually relies on experience to select the number of layers. The frequency of the initial reflection signal obtained by the anchor in acoustic wave detection is generally concentrated in the low-frequency part. At this time, a larger number of decomposition layers is required for more detailed analysis. Therefore, the selection of the number of decomposition layers in wavelet denoising under the traditional method is relatively small compared to the number of decomposition layers required for anchor quality defect detection, which may lead to poor denoising results for the initial reflection signal, making the acoustic wave reflection method unable to detect the anchor quality more accurately.

[0025] Therefore, this embodiment performs acoustic wave detection on any anchor rod to be detected to obtain an initial reflection signal. By analyzing the distribution characteristics of the signal in the initial reflection signal and the influence of the noise signal in the initial reflection signal, the optimal number of decomposition layers when performing wavelet denoising on the initial reflection signal is obtained, thereby improving the denoising effect of the initial reflection signal, so that the acoustic wave reflection method can more accurately detect anchor rod quality defects.

[0026] First, any anchor to be detected is recorded as a target anchor, and the target anchor is subjected to acoustic wave detection using an acoustic wave reflection method to obtain an initial reflection signal of the target anchor. The method of obtaining the initial reflection signal using the acoustic wave reflection method belongs to the prior art and will not be described in detail here.

[0027] Step S102: obtaining the signal complexity of the initial reflection signal according to the distribution characteristics of the signal in the initial reflection signal, obtaining the noise ratio of the initial reflection signal according to the influence of the noise signal in the initial reflection signal, and obtaining the decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal according to the signal complexity and noise ratio of the initial reflection signal.

[0028] After obtaining the initial reflection signal, the initial reflection signal is denoised. Since the number of decomposition layers selected in the traditional wavelet denoising is relatively small compared to the number of decomposition layers required for anchor quality defect detection, it may lead to poor denoising results for the initial reflection signal. Therefore, it is necessary to adjust the number of decomposition layers selected by the traditional wavelet decomposition according to the characteristics of the initial reflection signal itself, so as to obtain the optimal number of decomposition layers for wavelet denoising of the initial reflection signal and improve the denoising effect of the initial reflection signal.

[0029] When performing wavelet denoising on the initial reflection signal, the more complex the signal of the initial reflection signal is, the more decomposition layers the wavelet denoising needs to capture the subtle features in the initial reflection signal, thereby effectively removing the noise signal without losing the useful signal; the higher the component of the noise signal in the initial reflection signal, the more decomposition layers the wavelet denoising needs to increase to more effectively separate the noise signal from the useful signal and enhance the denoising effect.

[0030] Therefore, according to the signal complexity in the initial reflection signal and the proportion of noise in the initial reflection signal, the decomposition layer number adjustment coefficient for wavelet denoising of the initial reflection signal can be obtained, and then the decomposition layer number adjustment coefficient can be used to adjust the decomposition layer number for wavelet denoising of the initial reflection signal to obtain the optimal decomposition layer number.

[0031] Among them, according to the signal complexity in the initial reflection signal and the noise proportion in the initial reflection signal, the method for obtaining the decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal is as follows: (1) According to the distribution characteristics of the signal in the initial reflection signal, the signal complexity of the initial reflection signal is obtained.

[0032] Since the sample entropy can measure the complexity of the signal sequence, the larger the entropy value, the greater the complexity of the signal, and the bandwidth can reflect the range of signal frequency change. The larger the bandwidth, the more different frequency components the signal contains, which means the signal has a higher complexity. Therefore, the signal complexity of the initial reflection signal can be obtained based on the sample entropy and bandwidth of the initial reflection signal.

[0033] Specifically, performing wavelet denoising on the initial reflection signal to obtain a comparison signal, forming a comparison signal sequence from the comparison signal, obtaining sample entropy of the comparison signal sequence, and normalizing the sample entropy to obtain a first complexity of the initial reflection signal; Construct the spectrum diagram of the initial reflection signal, such as Figure 3 As shown, wherein the horizontal axis of the spectrum diagram is frequency, and the vertical axis is signal amplitude, the bandwidth of the initial reflection signal is obtained according to the spectrum diagram, and the bandwidth is normalized to obtain the second complexity of the initial reflection signal; A weighted sum is performed on the first complexity and the second complexity to obtain the signal complexity of the initial reflected signal.

[0034] In one embodiment, the calculation formula of the signal complexity of the initial reflection signal is:

[0035] in, is the signal complexity of the initial reflection signal; S is the sample entropy of the comparison signal sequence; B is the bandwidth of the initial reflection signal; norm() is the normalization function.

[0036] It should be noted that since the noise in the initial reflection signal may affect the calculation of the sample entropy, the sample entropy of the comparison signal sequence after wavelet denoising represents the signal complexity in the initial reflection signal. The larger the sample entropy of the comparison signal sequence, the higher the signal complexity of the initial reflection signal; the larger the bandwidth of the initial reflection signal, the higher the signal complexity of the initial reflection signal.

[0037] (2) According to the influence of the noise signal in the initial reflection signal, the noise ratio of the initial reflection signal is obtained.

[0038] Since power spectral density is usually used to analyze the spectral characteristics of a signal and reflect the distribution of frequency domain energy, a Fourier transform is performed on the initial reflected signal to obtain a power spectrum curve graph of the initial reflected signal, wherein the horizontal axis of the power spectrum curve graph is frequency and the vertical axis is power. The power spectrum density corresponding to each frequency in the spectrum graph is obtained in the power spectrum curve graph. The larger the power spectrum density corresponding to each frequency in the spectrum graph, the greater the energy corresponding to this frequency is. The maximum power spectrum density is obtained from the power spectrum density corresponding to each frequency in the spectrum graph, and the frequency corresponding to the maximum power spectrum density is recorded as the target frequency. All power spectrum densities except the maximum power spectrum density are combined into a power spectrum density sequence. Except for the target frequency, the power spectrum densities of the remaining frequencies should be small. However, if there is noise interference, the power spectrum densities of the remaining frequencies will be large. Therefore, the noise energy proportion of the initial reflected signal can be obtained according to the difference in power spectrum density corresponding to each frequency in the spectrum graph.

[0039] Among them, according to the difference of power spectrum density corresponding to each frequency in the spectrum diagram, the noise energy proportion of the initial reflection signal is obtained, including: For any power spectral density in the power spectral density sequence, obtain the absolute value of the frequency difference between the frequency corresponding to the any power spectral density and the target frequency, obtain the proportion of the absolute value of the frequency difference in the absolute values ​​of the frequency differences corresponding to all power spectral densities in the power spectral density sequence, and obtain the weight coefficient of the any power spectral density; Obtain a weight coefficient for each power spectral density in the power spectral density sequence, perform weighted summation on all power spectral densities in the power spectral density sequence to obtain a noise energy value, normalize the ratio of the noise energy value to the maximum power spectral density, and obtain the noise energy ratio of the initial reflection signal.

[0040] In one embodiment, the calculation formula for the noise energy ratio of the initial reflection signal is:

[0041] in, is the noise energy proportion of the initial reflection signal; is the frequency corresponding to the i-th power spectrum density in the power spectrum density sequence; f is the target frequency; is the i-th power spectral density in the power spectral density sequence; M is the number of power spectral densities in the power spectral density sequence; is the maximum power spectral density; norm() is the normalization function; is the absolute value symbol.

[0042] It should be noted that It is the weighted sum of all power spectrum densities in the power spectrum density sequence. In the spectrum diagram, the farther away the frequency is from the target frequency, the less likely it is to be the main frequency of the signal, that is, the greater the probability that the frequency is a noise component. At this time, a larger weight should be given to the frequency to highlight the influence of noise. Therefore, the proportion of the absolute value of the frequency difference between it and the target frequency in the absolute value of the frequency difference corresponding to all power spectrum densities in the power spectrum density sequence is taken as the weight coefficient of the corresponding power spectrum density. The larger it is, the greater the noise energy in the initial reflected signal, the greater the proportion of noise energy in the initial reflected signal, and the greater the proportion of noise in the initial reflected signal.

[0043] Furthermore, considering that when there are defects in the target anchor rod, the noise energy ratio of the initial reflection signal will be inaccurate, and the noise signal in the initial reflection signal will affect the average power of the initial reflection signal. The larger the noise ratio, the larger the average power of the initial reflection signal. Therefore, the average power of the initial reflection signal can be obtained according to the power spectrum curve, which is recorded as the average power of the initial reflection signal. Similarly, the power spectrum curve of the comparison signal is obtained, and the average power of the comparison signal is obtained according to the power spectrum curve of the comparison signal, which is recorded as the average power of the comparison signal. Then, the average power difference between the average power of the initial reflection signal and the average power of the comparison signal is calculated, and the ratio of the average power difference to the average power of the initial reflection signal is normalized to obtain the noise power ratio of the initial reflection signal.

[0044] In one implementation, the calculation formula for the noise power ratio of the initial reflected signal is:

[0045] in, is the noise power ratio of the initial reflected signal; P is the average power of the initial reflected signal; is the average power of the comparison signal; norm() is the normalization function.

[0046] It should be noted that is the average power difference between the initial reflected signal average power and the comparison signal average power, The larger it is, the greater the noise interference in the initial reflected signal, the greater the proportion of noise power in the initial reflected signal, and the greater the proportion of noise in the initial reflected signal.

[0047] Finally, the noise energy proportion and the noise power proportion of the initial reflected signal are combined to obtain the noise proportion of the initial reflected signal. Specifically, a weighted sum of the noise energy proportion and the noise power proportion is performed to obtain the noise proportion of the initial reflected signal.

[0048] In one embodiment, the calculation formula of the noise ratio of the initial reflection signal is:

[0049] in, is the noise proportion of the initial reflection signal; is the noise energy proportion of the initial reflection signal; is the noise power ratio of the initial reflected signal; is the weight of the noise energy proportion; is the weight of the noise power ratio.

[0050] It should be noted that since the noise energy ratio is more accurate than the noise power ratio, setting , There is no restriction here and it can be set according to the specific implementation scenario.

[0051] (3) Performing a weighted summation on the signal complexity and noise proportion of the initial reflection signal to obtain a decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal.

[0052] In one embodiment, the calculation formula of the decomposition layer adjustment coefficient is:

[0053] in, is the adjustment factor for the number of decomposition layers; is the signal complexity of the initial reflected signal; is the noise proportion of the initial reflection signal; is the weight of signal complexity; is the weight of the noise ratio.

[0054] It should be noted that the signal complexity and noise ratio of the initial reflection signal are equally important for the selection of the number of decomposition layers when performing wavelet denoising on the initial reflection signal. , there is no restriction here and it can be set according to the specific implementation scenario; the greater the signal complexity of the initial reflection signal, the more decomposition layers the wavelet denoising needs to capture the subtle features in the initial reflection signal, and the larger the adjustment coefficient of the decomposition layers; the greater the noise proportion of the initial reflection signal, the greater the wavelet denoising needs to increase the decomposition layers to more effectively separate the noise signal and the useful signal, and the larger the adjustment coefficient of the decomposition layers.

[0055] At this point, the decomposition layer number adjustment coefficient for wavelet denoising of the initial reflection signal is obtained.

[0056] Step S103, obtaining the number of decomposition layers when performing wavelet denoising on the initial reflection signal, adjusting the number of decomposition layers according to the decomposition layer number adjustment coefficient, and obtaining the optimal number of decomposition layers for performing wavelet denoising on the initial reflection signal.

[0057] Since the size of the decomposition layer has a great influence on the denoising result during wavelet denoising, a small decomposition layer may not be able to capture and remove the noise components in the signal, and a large decomposition layer may cause signal distortion. Therefore, it is necessary to use the obtained decomposition layer adjustment coefficient to adjust the decomposition layer of the initial reflection signal during wavelet denoising to obtain the optimal decomposition layer to improve the effect of wavelet denoising. Since a higher decomposition layer can process the initial reflection signal at a finer scale and enhance the denoising effect, the higher the complexity of the initial reflection signal, the more it contains multiple frequency components, and a more detailed decomposition is required to effectively remove the noise signal without losing the useful signal; the larger the noise ratio of the initial reflection signal, the more decomposition layers need to be increased to more effectively separate the noise signal from the useful signal.

[0058] Among them, the method of using the obtained decomposition layer adjustment coefficient to adjust the decomposition layer number when performing wavelet denoising on the initial reflection signal to obtain the optimal decomposition layer number is as follows: According to the experimental results, the threshold of the decomposition layer adjustment coefficient is set to 0.6. When the decomposition layer adjustment coefficient is greater than 0.6, it means that the signal complexity in the initial reflection signal is large and the noise accounts for a high proportion. At this time, the decomposition layer needs to be increased to more effectively separate the noise signal and the useful signal and enhance the denoising effect. Therefore, the decomposition layer adjustment coefficient is used as the independent variable of the Sigmoid function to obtain the Sigmoid function value, and the addition result of the constant 1 and the Sigmoid function value is obtained to obtain the expansion coefficient of the decomposition layer. The product of the expansion coefficient and the decomposition layer number is rounded up to obtain the optimal decomposition layer number for wavelet denoising of the initial reflection signal.

[0059] In one embodiment, the calculation formula for the optimal number of decomposition layers is:

[0060] in, is the optimal number of decomposition layers; is the decomposition layer adjustment coefficient; L is the decomposition layer number when performing wavelet denoising on the initial reflection signal; Sigmoid() is an S-type function, also known as an S-type growth curve, which can map the output result to (0, 1); is the rounding symbol; 1 is a constant.

[0061] It should be noted that the larger the decomposition layer adjustment coefficient is, the greater the signal complexity and the higher the noise ratio in the initial reflection signal is, and the more it is necessary to adjust the decomposition layer number when performing wavelet denoising on the initial reflection signal, and the larger the optimal decomposition layer number is.

[0062] When the decomposition layer adjustment coefficient is less than or equal to 0.6, it means that the signal complexity and noise ratio in the initial reflection signal are relatively low. At this time, the decomposition layer number of wavelet denoising obtained by the traditional method is sufficient to denoise the initial reflection signal. Further adjustment of the decomposition layer number may increase the computational complexity. Therefore, no adjustment is made at this time, and the decomposition layer number is used as the optimal decomposition layer number for wavelet denoising of the initial reflection signal.

[0063] So far, the optimal number of decomposition layers for wavelet denoising of the initial reflection signal is obtained.

[0064] Step S104, using the optimal decomposition layer number as the decomposition layer number of wavelet denoising, performing wavelet denoising on the initial reflection signal to obtain a denoised signal, and performing quality defect detection on any anchor rod to be inspected according to the denoised signal.

[0065] After obtaining the optimal decomposition layer number for wavelet denoising of the initial reflection signal, the optimal decomposition layer number is used as the decomposition layer number for wavelet denoising of the initial reflection signal, and the initial reflection signal is subjected to wavelet denoising to obtain a denoised signal. Quality defect detection of the target anchor rod is performed based on the denoised signal, wherein wavelet denoising and quality defect detection of the target anchor rod belong to the existing technology and will not be repeated here.

[0066] In this embodiment, for any anchor rod to be inspected, acoustic wave inspection is performed on the any anchor rod to be inspected to obtain an initial reflection signal; based on the distribution characteristics of the signal in the initial reflection signal, the signal complexity of the initial reflection signal is obtained; based on the influence of the noise signal in the initial reflection signal, the noise proportion of the initial reflection signal is obtained; based on the signal complexity and noise proportion of the initial reflection signal, an adjustment coefficient of the number of decomposition layers when performing wavelet denoising on the initial reflection signal is obtained; the number of decomposition layers when performing wavelet denoising on the initial reflection signal is obtained, and the number of decomposition layers is adjusted according to the decomposition layer adjustment coefficient to obtain the optimal number of decomposition layers for wavelet denoising on the initial reflection signal; the optimal number of decomposition layers is used as the number of decomposition layers for wavelet denoising, wavelet denoising is performed on the initial reflection signal to obtain a denoised signal, and quality defect detection is performed on any anchor rod to be inspected according to the denoised signal. Among them, according to the distribution characteristics of the signal in the obtained initial reflection signal and the influence of the noise signal, the decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal is obtained, and then the decomposition layer number is adjusted according to the decomposition layer number adjustment coefficient, and the optimal decomposition layer number for wavelet denoising on the initial reflection signal is accurately obtained, which reduces the poor noise removal effect caused by the traditional method of relying on experience to select the wavelet denoising decomposition layer number, and the optimal decomposition layer number is more dependent on the characteristics of the initial reflection signal itself, so that the acoustic wave reflection method can more accurately detect anchor quality defects, thereby improving the accuracy of anchor quality defect detection.

[0067] The above embodiments are only used to illustrate the technical solutions of the present invention, rather than to limit the same. Although the present invention has been described in detail with reference to the aforementioned embodiments, those skilled in the art should understand that the technical solutions described in the aforementioned embodiments may still be modified, or some of the technical features may be replaced by equivalents. Such modifications or replacements do not deviate the essence of the corresponding technical solutions from the spirit and scope of the technical solutions of the embodiments of the present invention, and should be included in the protection scope of the present invention.

Claims

1. A method for detecting anchor quality defects, characterized in that: The method comprises: For any anchor rod to be detected, performing acoustic wave detection on the anchor rod to be detected to obtain an initial reflection signal; According to the distribution characteristics of the signal in the initial reflection signal, the signal complexity of the initial reflection signal is obtained, according to the influence degree of the noise signal in the initial reflection signal, the noise proportion of the initial reflection signal is obtained, and according to the signal complexity and noise proportion of the initial reflection signal, the decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal is obtained; Obtaining the number of decomposition layers when performing wavelet denoising on the initial reflection signal, and adjusting the number of decomposition layers according to the decomposition layer number adjustment coefficient to obtain the optimal number of decomposition layers for performing wavelet denoising on the initial reflection signal; The optimal decomposition layer number is used as the decomposition layer number of wavelet denoising, wavelet denoising is performed on the initial reflection signal to obtain a denoised signal, and quality defect detection is performed on any anchor rod to be inspected based on the denoised signal.

2. The anchor bolt quality defect detection method according to claim 1, characterized in that: The acquiring the signal complexity of the initial reflected signal according to the distribution characteristics of the signal in the initial reflected signal comprises: Performing wavelet denoising on the initial reflection signal to obtain a comparison signal, composing the comparison signal into a comparison signal sequence, obtaining sample entropy of the comparison signal sequence, and normalizing the sample entropy to obtain a first complexity of the initial reflection signal; Constructing a frequency spectrum of the initial reflection signal, wherein the abscissa of the frequency spectrum is the frequency and the ordinate is the signal amplitude, obtaining the bandwidth of the initial reflection signal according to the frequency spectrum, and performing normalization processing on the bandwidth to obtain the second complexity of the initial reflection signal; A weighted sum is performed on the first complexity and the second complexity to obtain the signal complexity of the initial reflected signal.

3. The method for detecting anchor quality defects according to claim 2, characterized in that: The obtaining, according to the influence degree of the noise signal in the initial reflection signal, the noise proportion of the initial reflection signal comprises: Performing Fourier transform on the initial reflected signal to obtain a power spectrum curve graph of the initial reflected signal, wherein the abscissa of the power spectrum curve graph is frequency and the ordinate is power; Obtaining the power spectrum density corresponding to each frequency in the spectrum graph in the power spectrum curve graph, and obtaining the noise energy proportion of the initial reflected signal according to the difference in the power spectrum density corresponding to each frequency in the spectrum graph; According to the power spectrum curve, the average power of the initial reflected signal is obtained, which is recorded as the initial reflected signal average power; the power spectrum curve of the comparison signal is obtained; according to the power spectrum curve of the comparison signal, the average power of the comparison signal is obtained, which is recorded as the comparison signal average power; according to the difference between the initial reflected signal average power and the comparison signal average power, the noise power ratio of the initial reflected signal is obtained; A weighted sum is performed on the noise energy proportion and the noise power proportion to obtain the noise proportion of the initial reflected signal.

4. A method for detecting anchor quality defects according to claim 3, characterized in that: The obtaining the noise energy proportion of the initial reflected signal according to the difference of the power spectrum density corresponding to each frequency in the spectrum diagram includes: Obtaining a maximum power spectrum density from the power spectrum densities corresponding to each frequency in the spectrum graph, and composing a power spectrum density sequence from all power spectrum densities except the maximum power spectrum density; Record the frequency corresponding to the maximum power spectrum density as the target frequency, and for any power spectrum density in the power spectrum density sequence, obtain the absolute value of the frequency difference between the frequency corresponding to any power spectrum density and the target frequency, obtain the proportion of the absolute value of the frequency difference in the absolute values ​​of the frequency differences corresponding to all power spectrum densities in the power spectrum density sequence, and obtain the weight coefficient of any power spectrum density; Obtain a weight coefficient for each power spectral density in the power spectral density sequence, perform weighted summation on all power spectral densities in the power spectral density sequence to obtain a noise energy value, normalize the ratio of the noise energy value to the maximum power spectral density, and obtain the noise energy ratio of the initial reflection signal.

5. The anchor bolt quality defect detection method according to claim 3, characterized in that: The obtaining, according to the difference between the average power of the initial reflected signal and the average power of the comparison signal, the noise power ratio of the initial reflected signal comprises: The average power difference between the average power of the initial reflected signal and the average power of the comparison signal is calculated, and the ratio of the average power difference to the average power of the initial reflected signal is normalized to obtain the noise power ratio of the initial reflected signal.

6. The anchor bolt quality defect detection method according to claim 1, characterized in that: The step of obtaining a decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal according to the signal complexity and noise proportion of the initial reflection signal comprises: A weighted sum is taken for the signal complexity and noise proportion of the initial reflection signal to obtain a decomposition layer number adjustment coefficient when performing wavelet denoising on the initial reflection signal.

7. The anchor bolt quality defect detection method according to claim 1, characterized in that: The step of adjusting the number of decomposition layers according to the decomposition layer number adjustment coefficient to obtain an optimal number of decomposition layers for performing wavelet denoising on the initial reflection signal comprises: When the decomposition layer number adjustment coefficient is greater than a preset decomposition layer number adjustment coefficient threshold, the decomposition layer number adjustment coefficient is used as an independent variable of a Sigmoid function to obtain a Sigmoid function value; The addition result of the constant 1 and the Sigmoid function value is obtained to obtain the expansion coefficient of the decomposition layer number, and the product of the expansion coefficient and the decomposition layer number is rounded up to obtain the optimal decomposition layer number for wavelet denoising of the initial reflection signal.

8. The anchor bolt quality defect detection method according to claim 7, characterized in that: The step of adjusting the number of decomposition layers according to the decomposition layer number adjustment coefficient to obtain the optimal number of decomposition layers for wavelet denoising of the initial reflection signal further includes: When the decomposition layer number adjustment coefficient is less than or equal to a preset decomposition layer number adjustment coefficient threshold, the decomposition layer number is used as the optimal decomposition layer number for wavelet denoising of the initial reflection signal.

Images