Cast-in-place pile quality online monitoring method and system based on ultrasonic array

Abstract

The invention provides a cast-in-place pile quality online monitoring method and system based on an ultrasonic array, and the method comprises the steps: transmitting and receiving ultrasonic waves through the ultrasonic array arranged in a cast-in-place pile, and obtaining a multi-channel sound wave signal; performing generalized S transformation processing on the multichannel sound wave signals to obtain a time-frequency matrix, and calculating a cross correlation coefficient between each channel signal and the defect-free reference signal; an amplitude matrix and an instantaneous phase gradient matrix are calculated based on the time-frequency matrix, neighborhood amplitude energy and a phase consistency factor are fused in a time-frequency domain, and a defect energy indication diagram of the pile body section is generated; calculating the energy entropy of the defect energy indication diagram in a full frequency band; and extracting an actually measured energy arrival time sequence, comparing the actually measured energy arrival time sequence with a theoretical arrival time sequence to construct a space distortion vector, and setting a threshold value by combining energy entropy to judge defects.

Description

Technical Field

[0001] This application belongs to the field of monitoring, and in particular relates to an online monitoring method and system for the quality of cast-in-place piles based on ultrasonic arrays. Background Technology

[0002] The integrity and bearing capacity of cast-in-place piles are crucial to the safety and stability of the entire structure. Ultrasonic testing technology, especially the acoustic transmission method, analyzes the changes in acoustic parameters after ultrasonic waves pass through the concrete of the pile to determine the presence of defects such as diameter reduction, mud inclusion, segregation, and voids. The complex on-site construction environment often results in ultrasonic signals mixed with strong background noise and random interference, and the signals themselves exhibit typical non-stationary and nonlinear characteristics due to propagation in a non-uniform medium. Signal processing methods, such as Fourier transform, cannot simultaneously handle localized analysis in both the time and frequency domains. While wavelet transform offers some improvement, its time-frequency resolution is globally fixed once selected, making it difficult to adapt to signal variations. The standard S-transform combines the advantages of short-time Fourier transform and continuous wavelet transform, providing excellent time-frequency focusing. However, the symmetrical Gaussian window function used in processing impact ultrasonic signals with steep leading edges or rapid attenuation cannot accurately match the local energy distribution of the signal, easily leading to time-frequency ambiguity and energy leakage. Defect identification methods often rely on acoustic features such as acoustic time or amplitude attenuation, resulting in insufficient information utilization. When defect morphology is complex or the signal-to-noise ratio is low, missed or false detections are prone to occur. Furthermore, the threshold setting for defect discrimination lacks the ability to adjust for specific pile parameters and overall signal quality, leading to a need to improve the objectivity and reliability of the detection results. Therefore, developing an online monitoring method that can improve time-frequency resolution, integrate multi-dimensional feature information, and achieve intelligent defect identification, based on the characteristics of ultrasonic testing signals for cast-in-place piles, is a pressing technical challenge in this field. Summary of the Invention

[0003] This invention proposes an online monitoring method for the quality of cast-in-place piles based on ultrasonic arrays, comprising the following steps: An ultrasonic array arranged inside the cast-in-place pile is used to transmit and receive ultrasonic waves to obtain multi-channel acoustic signals. The generalized S-transform is performed on each channel signal in the multi-channel acoustic signal to obtain their respective time-frequency matrices, and the cross-correlation coefficient between the channel signal and the preset defect-free reference signal is calculated. Based on the time-frequency matrix, the amplitude matrix and the instantaneous phase gradient matrix are calculated respectively; in the time-frequency domain, the neighborhood amplitude energy and neighborhood phase consistency factor of each point are fused to generate a defect energy indication map of the pile section, wherein the weighting coefficient of the fusion is determined by the cross-correlation coefficient calculated above; the energy entropy of the defect energy indication map in the whole frequency band is calculated. In the defect energy indication diagram, the peak energy times of each channel within a preset frequency band are extracted along the frequency axis to form a measured energy arrival time series; the difference between the measured energy arrival time series and the theoretical arrival time series is calculated to construct a spatial distortion vector; a first threshold and a second threshold are set that are in a preset functional relationship with the reciprocal of the energy entropy; when the magnitude of the spatial distortion vector exceeds the first threshold and the number of components exceeding the second threshold reaches a preset number, it is determined that there is a defect in the corresponding pile section.

[0004] Optionally, the generalized S-transform employs an asymmetric Gaussian window function, the skewness of which has a preset nonlinear mapping relationship with the local energy gradient of the signal within the time window. The frequency adjustment factor of the generalized S-transform is selected based on the initial signal-to-noise ratio of the channel signal, including: The skewness of the window function Through formula Determine, where G(t,f) is the local energy gradient of the signal within the time window. and The frequency adjustment factor p of the generalized S-transform is a preset constant; the frequency adjustment factor p of the generalized S-transform is expressed by the formula... Confirmed, among which Let be the initial signal-to-noise ratio of the channel signal, and k and b be preset constants.

[0005] Optionally, for each point (t,f) in the time-frequency matrix, a 3×3 rectangular neighborhood is selected. Calculate the sum of squared magnitudes of all 9 points in the neighborhood, which is taken as the neighborhood magnitude energy E(t,f) of the point; Calculate the magnitude of the vector sum of the instantaneous phase gradients of all 9 points in the neighborhood, and use it as the neighborhood phase consistency factor P(t,f) of the point.

[0006] Optionally, the weighting coefficients of the fusion are determined by the cross-correlation coefficients calculated above, including: Let the cross-correlation coefficient be R, then the weighting coefficient of the neighborhood magnitude energy... The weighting coefficient of the neighborhood phase consistency factor ; Normalize all calculated neighborhood magnitude energies E(t,f) and neighborhood phase coherence factors P(t,f) to obtain normalized neighborhood magnitude energies. Phase consistency factor with neighborhood ; The energy value D(t,f) of point (t,f) in the defect energy indicator map is obtained through a weighted fusion formula. Calculated.

[0007] Optionally, calculating the energy entropy of the defect energy indicator map across the entire frequency band includes: Defect energy indicator map The total energy is obtained by summing all the energy values. And calculate the probability of each time frequency point. ; The energy entropy is calculated using the Shannon entropy formula.

[0008] Optionally, the step of calculating the difference between the measured energy arrival time series and the theoretical arrival time series to construct a spatial distortion vector includes: The theoretical arrival time series The i-th component is determined by the linear geometric distance between the ultrasonic transmitting and receiving probes corresponding to the i-th channel. and the preset sound velocity of the pile concrete Through formula Calculated; Measured energy arrival time series The components and theoretical arrival time series Subtracting the corresponding components yields the spatial distortion vector. .

[0009] Optionally, the setting of the first threshold and the second threshold, which are related to the reciprocal of the energy entropy by a preset function, includes: Energy entropy is H, first threshold From the linear formula Calculate the second threshold. From the linear formula Calculation, where This is a constant obtained by calibrating standard defect-free pile experimental data.

[0010] Optionally, determining that the corresponding pile section has a defect when the magnitude of the spatial distortion vector exceeds a first threshold and the number of components exceeding a second threshold reaches a preset number includes: The square root of the sum of the squares of all components of the spatial distortion vector is taken as the magnitude of the spatial distortion vector. When the magnitude of the spatial distortion vector is greater than the first threshold, and the number of components in the vector whose absolute value is greater than the second threshold reaches the preset number of 3 or more, a defect is determined to exist.

[0011] Furthermore, the present invention also relates to an online monitoring system for the quality of cast-in-place piles based on an ultrasonic array, comprising the following modules: The acquisition module is used to transmit and receive ultrasonic waves using an ultrasonic array arranged in the cast-in-place pile, and to acquire multi-channel acoustic signals. The selection module is used to perform generalized S-transform processing on each channel signal in the multi-channel acoustic signal to obtain their respective time-frequency matrices, and to calculate the cross-correlation coefficient between the channel signal and the preset defect-free reference signal. The calculation module is used to calculate the amplitude matrix and the instantaneous phase gradient matrix based on the time-frequency matrix; in the time-frequency domain, it fuses the neighborhood amplitude energy and neighborhood phase consistency factor of each point to generate a defect energy indication map of the pile cross section, wherein the weighting coefficient of the fusion is determined by the cross-correlation coefficient calculated above; and calculates the energy entropy of the defect energy indication map in the whole frequency band. The determination module is used to extract the peak energy times of each channel within a preset frequency band along the frequency axis in the defect energy indication diagram to form a measured energy arrival time series; calculate the difference between the measured energy arrival time series and the theoretical arrival time series to construct a spatial distortion vector; set a first threshold and a second threshold that are in a preset functional relationship with the reciprocal of the energy entropy; when the magnitude of the spatial distortion vector exceeds the first threshold and the number of components exceeding the second threshold reaches a preset number, it is determined that there is a defect in the corresponding pile section.

[0012] Preferably, the generalized S-transform employs an asymmetric Gaussian window function, the skewness of which has a preset nonlinear mapping relationship with the local energy gradient of the signal within the time window, and the frequency adjustment factor of the generalized S-transform is selected based on the initial signal-to-noise ratio of the channel signal, including: The skewness of the window function Through formula Determine, where G(t,f) is the local energy gradient of the signal within the time window. and The frequency adjustment factor p of the generalized S-transform is a preset constant; the frequency adjustment factor p of the generalized S-transform is expressed by the formula... Confirmed, among which Let be the initial signal-to-noise ratio of the channel signal, and k and b be preset constants.

[0013] Preferably, for each point (t,f) in the time-frequency matrix, a 3×3 rectangular neighborhood is selected. Calculate the sum of squared magnitudes of all 9 points in the neighborhood, which is taken as the neighborhood magnitude energy E(t,f) of the point; Calculate the magnitude of the vector sum of the instantaneous phase gradients of all 9 points in the neighborhood, and use it as the neighborhood phase consistency factor P(t,f) of the point.

[0014] Preferably, the fusion weighting coefficients are determined by the cross-correlation coefficients calculated above, including: Let the cross-correlation coefficient be R, then the weighting coefficient of the neighborhood magnitude energy... The weighting coefficient of the neighborhood phase consistency factor ; Normalize all calculated neighborhood magnitude energies E(t,f) and neighborhood phase coherence factors P(t,f) to obtain normalized neighborhood magnitude energies. Phase consistency factor with neighborhood ; The energy value D(t,f) of point (t,f) in the defect energy indicator map is obtained through a weighted fusion formula. Calculated.

[0015] Preferably, calculating the energy entropy of the defect energy indicator map across the entire frequency band includes: Defect energy indicator map The total energy is obtained by summing all the energy values. And calculate the probability of each time frequency point. ; The energy entropy is calculated using the Shannon entropy formula.

[0016] Preferably, the step of calculating the difference between the measured energy arrival time series and the theoretical arrival time series to construct a spatial distortion vector includes: The theoretical arrival time series The i-th component is determined by the linear geometric distance between the ultrasonic transmitting and receiving probes corresponding to the i-th channel. and the preset sound velocity of the pile concrete Through formula Calculated; Measured energy arrival time series The components and theoretical arrival time series Subtracting the corresponding components yields the spatial distortion vector. .

[0017] Preferably, the setting of the first threshold and the second threshold, which are preset functionally related to the reciprocal of the energy entropy, includes: Energy entropy is H, first threshold From the linear formula Calculate the second threshold. From the linear formula Calculation, where This is a constant obtained by calibrating standard defect-free pile experimental data.

[0018] Preferably, determining that the corresponding pile section has a defect when the magnitude of the spatial distortion vector exceeds a first threshold and the number of components exceeding a second threshold reaches a preset number includes: The square root of the sum of the squares of all components of the spatial distortion vector is taken as the magnitude of the spatial distortion vector. When the magnitude of the spatial distortion vector is greater than the first threshold, and the number of components in the vector whose absolute value is greater than the second threshold reaches the preset number of 3 or more, a defect is determined to exist.

[0019] This invention improves the time-frequency resolution and noise immunity of the generalized S-transform by employing an asymmetric Gaussian window function matched to the local energy gradient of the signal and a frequency adjustment factor selected based on the initial signal-to-noise ratio, thus obtaining a more accurate signal time-frequency distribution. By fusing amplitude energy and phase consistency information in the time-frequency domain and using the cross-correlation coefficient of the signal to determine the fusion weight, it can more sensitively reflect the anomalies inside the pile body, generating a defect indication map with higher contrast. A defect discrimination criterion is established that not only considers the spatial distortion of the wavefield arrival time but also uses the energy entropy representing the overall complexity of the signal to construct the discrimination threshold. Joint judgment is performed under dual conditions, thereby establishing an intrinsic link between the overall signal quality and local criteria, making the defect identification results more objective and reliable, and reducing missed and false judgments caused by noise interference. Attached Figure Description

[0020] Figure 1 A flowchart of the first embodiment; Figure 2 This is a schematic diagram illustrating the relationship between the frequency adjustment factor and the signal-to-noise ratio. Figure 3 This is a schematic diagram of a spatially distorted vector; Figure 4 This is a schematic diagram illustrating the relationship between the threshold and energy entropy. Detailed Implementation

[0021] Many specific details are set forth in the following description to provide a full understanding of this specification. However, this specification can be implemented in many other ways than those described herein, and those skilled in the art can make similar extensions without departing from the spirit of this specification. Therefore, this specification is not limited to the specific implementations disclosed below.

[0022] The terminology used in one or more embodiments of this specification is for the purpose of describing particular embodiments only and is not intended to be limiting of the one or more embodiments of this specification. The singular forms “a,” “described,” and “the” as used in one or more embodiments of this specification and the appended claims are also intended to include the plural forms unless the context clearly indicates otherwise. It should also be understood that the term “and / or” as used in one or more embodiments of this specification refers to and includes any or all possible combinations of one or more associated listed items.

[0023] It should be understood that although the terms first, second, etc., may be used to describe various information in one or more embodiments of this specification, such information should not be limited to these terms. These terms are only used to distinguish information of the same type from one another. For example, first may also be referred to as second without departing from the scope of one or more embodiments of this specification, and similarly, second may also be referred to as first. Depending on the context, the word "if" as used herein may be interpreted as "when," "when," or "in response to a determination."

[0024] In the first embodiment, the present invention proposes an online monitoring method for the quality of cast-in-place piles based on an ultrasonic array, such as... Figure 1 This includes the following steps: S1, using an ultrasonic array arranged in the cast-in-place pile to transmit and receive ultrasonic waves to obtain multi-channel acoustic signals; Multiple ultrasonic transducers are pre-embedded within a reinforcing cage along the axial and circumferential directions of the pile, forming a transmitting and receiving array. Controlled by a data acquisition system, one transducer is sequentially excited as the transmitting probe, emitting a pulse signal with a center frequency of 50kHz. All other transducers in the array synchronously receive the signal as receiving probes. This process is repeated until all transducers have been excited once as transmitting probes. The received signal is then converted from analog to digital using a multi-channel data acquisition card at a sampling rate of 1MHz, obtaining a series of discrete time-series signals, forming a multi-channel acoustic signal data matrix containing all transmitting and receiving paths.

[0025] S2, Perform generalized S-transform processing on each channel signal in the multi-channel acoustic signal to obtain their respective time-frequency matrices, and calculate the cross-correlation coefficient between the channel signal and the preset defect-free reference signal; Specifically, the pure noise portion of the initial signal segment and the signal portion after the first wave arrives are extracted from each channel, and their average energy is calculated separately. The ratio of these two values ​​is used as the initial signal-to-noise ratio (SNR). The frequency adjustment factor is calculated based on the initial SNR; preferably, the two are positively correlated. For each time window of the S-transform, the difference between the energy of the first half and the energy of the second half of the signal within the window is calculated as the local energy gradient G. The skewness γ of the asymmetric Gaussian window is mapped to G through the hyperbolic tangent function, i.e. The constant c is used to control the sensitivity of the mapping. Substituting the asymmetric Gaussian window function with skewness γ and frequency adjustment factor into the standard S-transform integral formula, point-by-point time-frequency analysis is performed on the signal to obtain the complex time-frequency matrix. Normalized cross-correlation is performed between the current channel signal and the pre-stored ideal pile model signal from the same location in the database. The maximum value of the correlation function is taken as the cross-correlation coefficient, such as... Figure 2 .

[0026] In another optional embodiment, the generalized S-transform employs an asymmetric Gaussian window function, the skewness of which has a preset nonlinear mapping relationship with the local energy gradient of the signal within the time window. The frequency adjustment factor of the generalized S-transform is selected based on the initial signal-to-noise ratio of the channel signal, including: The skewness of the window function Through formula Determine, where G(t,f) is the local energy gradient of the signal within the time window. and The frequency adjustment factor p of the generalized S-transform is a preset constant; the frequency adjustment factor p of the generalized S-transform is expressed by the formula... Confirmed, among which Let be the initial signal-to-noise ratio of the channel signal, and k and b be preset constants.

[0027] The S-transform effect is optimized by adjusting the shape and frequency resolution of the window function, specifically by adjusting the skewness γ of the window function. For example, at a given time-frequency point, if the signal energy increases sharply, the calculated local energy gradient G(t,f) will be a positive value, say 0.8. If the preset constants α and β are empirically set to 2 and 1.5 respectively, then the skewness γ will be approximately 1.82. A positive skewness will cause the Gaussian window function to tilt in the direction of energy growth, thus more accurately detecting the instantaneous characteristics of the signal.

[0028] The frequency adjustment factor p is selected. For example, after receiving a signal from one channel, the initial signal-to-noise ratio (SNR) is calculated. Assume the SNR of a clear signal is 30. Preset constants k and b are experimentally calibrated and set to 0.05 and 1, respectively. Then the value of the frequency adjustment factor p is 2.5. If the other channel signal has higher noise and an SNR of only 10, then the value of p is calculated to be 1.5.

[0029] S3, Based on the time-frequency matrix, calculate the amplitude matrix and the instantaneous phase gradient matrix respectively; In the time-frequency domain, fuse the neighborhood amplitude energy and neighborhood phase consistency factor of each point to generate a defect energy indication map of the pile section, wherein the weighting coefficient of the fusion is determined by the cross-correlation coefficient calculated above; Calculate the energy entropy of the defect energy indication map in the whole frequency band; Specifically, the magnitude matrix is ​​obtained by taking the modulus of each element in the complex time-frequency matrix. The phase angle of each element in the complex time-frequency matrix is ​​calculated to obtain the phase matrix. Then, a difference operation is performed on the phase matrix along the time axis to approximate the instantaneous phase gradient matrix. For any point in the time-frequency matrix, a 5×5 pixel region is defined as the neighborhood. The sum of the squares of the magnitudes of all points within this neighborhood is calculated as the neighborhood magnitude energy. The instantaneous phase gradient of each point within the neighborhood is considered as a two-dimensional vector. The ratio of the magnitude of the sum of these vectors to the sum of the magnitudes of the individual vectors is calculated as the neighborhood phase consistency factor, which ranges from 0 to 1. Based on the cross-correlation coefficient ρ obtained above, the weighting coefficient w = 0.5 + ρ is calculated. In one embodiment, the value of each point on the defect energy indicator map is equal to w × neighborhood magnitude energy plus (1 - w) × (1 - neighborhood phase consistency factor).

[0030] Integrating or summing the generated defect energy indicator map along the time axis yields a one-dimensional energy distribution sequence E(f) over frequency. This sequence is then normalized by dividing the energy value at each point by the total energy, resulting in p(f). Using the Shannon entropy formula, the logarithm of the energy entropy H = -p(f) × p(f) is summed over all frequency points.

[0031] In an alternative embodiment, to extract features from the time-frequency matrix that better represent the stability of the signal structure, For each point (t,f) in the time-frequency matrix, select a 3×3 rectangular neighborhood around it; Calculate the sum of squared magnitudes of all 9 points in the neighborhood, which is taken as the neighborhood magnitude energy E(t,f) of the point; Calculate the magnitude of the vector sum of the instantaneous phase gradients of all 9 points in the neighborhood, and use it as the neighborhood phase consistency factor P(t,f) of the point.

[0032] Specifically, for any center point in the time-frequency matrix, a nine-square grid region including the center point is determined. The neighborhood amplitude energy is calculated. Assuming the amplitudes of the nine points are 1.1, 1.2, 1.1, 1.5, 2.0, 1.6, 1.0, 1.1, and 1.0 respectively, the neighborhood amplitude energy E(t,f) of the center point is the sum of the squares of these nine values. The calculated result reflects the degree of energy concentration in this local area. A higher energy value indicates the possible presence of a stable signal wavefront.

[0033] To calculate the neighborhood phase coherence factor, for each point within the aforementioned nine-grid area, calculate the instantaneous phase and solve for the gradient of this phase in the time and frequency directions, obtaining a two-dimensional gradient vector. Nine points correspond to nine two-dimensional vectors. Summing these nine vectors yields a resultant vector. The length, or magnitude, of this resultant vector is the neighborhood phase coherence factor P(t,f) of the center point. If the signal phase changes at all points in the neighborhood follow the same trend, for example, all pointing in the direction of signal propagation, then the gradient vectors are similar in direction, and the magnitude of the vector sum is large. Conversely, if there is noise or signal reflection interference in the neighborhood, the phase becomes disordered, the gradient vectors are in different directions, and the vector sum becomes very small due to mutual cancellation.

[0034] In an optional embodiment, the fusion weight coefficients are determined by the cross-correlation coefficients calculated above, including: Let the cross-correlation coefficient be R, then the weighting coefficient of the neighborhood magnitude energy... The weighting coefficient of the neighborhood phase consistency factor ; Normalize all calculated neighborhood magnitude energies E(t,f) and neighborhood phase coherence factors P(t,f) to obtain normalized neighborhood magnitude energies. Phase consistency factor with neighborhood ; The energy value D(t,f) of point (t,f) in the defect energy indicator map is obtained through a weighted fusion formula. Calculated.

[0035] Weighted fusion combines amplitude and phase information. Weights are determined based on signal quality. For example, a cross-correlation coefficient R between the received signal and the standard signal calculated above is 0.8, indicating good signal quality and more reliable phase information. Therefore, the weight of the neighborhood phase coherence factor is... That is, 0.8, and the weight of the neighborhood amplitude energy. The value is 0.2. If the signal quality of the other channel is poor, and the R value is 0.3, then... It is 0.3. The value is 0.7, at which point the focus is more on energy information.

[0036] The neighborhood amplitude energy E(t,f) and neighborhood phase consistency factor P(t,f) of all time-frequency points are normalized to unify their numerical range to between 0 and 1, facilitating weighting. For example, the maximum and minimum values ​​of all E(t,f) values ​​in the entire time-frequency matrix are found, and the linear stretching formula is applied to normalize E(t,f) at each point. The same process is performed on P(t,f). For a given time-frequency point, assume the normalized energy... The normalized phase factor is 0.9. The value is 0.95, and the weight is... It is 0.2. If the value is 0.8, then the energy value D(t,f) of that point in the defect energy indicator map is 0.94. This process is repeated for all time-frequency points to generate a complete defect energy indicator map.

[0037] In an optional embodiment, calculating the energy entropy of the defect energy indication map across the entire frequency band includes: Defect energy indicator map The total energy is obtained by summing all the energy values. And calculate the probability of each time frequency point. ; The energy entropy is calculated using the Shannon entropy formula.

[0038] The defect energy indicator map D(t,f) is treated as a two-dimensional probability distribution, and its information entropy is calculated to represent the degree of disorder or uncertainty in the energy distribution. The total energy S is obtained by summing the energy values ​​of all pixels in the map. For example, the sum of the energy values ​​of all 16 points in a simplified 4x4 indicator map is 500. The energy percentage of each point is calculated, i.e., the probability. If the energy value of a point (i,j) is 10, then the probability is... The value is 0.02. The process is performed on all 16 points, resulting in 16 probability values, the sum of which is 1.

[0039] After obtaining the probability distribution, apply the Shannon entropy formula. Calculate the energy entropy H. Sum the calculation results for all 16 points; the total sum is the energy entropy of the indicator map. If the energy is highly concentrated at a few points, then the probability of most points being concentrated in a few areas is lower. The probability of sound scattering will be very small, but a few points will have a high probability, resulting in a lower calculated energy entropy H value, which usually corresponds to a healthy pile. Conversely, if the energy is scattered across multiple areas in the diagram, the probability values ​​at each point will be more even, resulting in a higher calculated energy entropy H value, suggesting the possible presence of defects causing sound wave scattering.

[0040] S4. In the defect energy indication diagram, extract the peak energy time of each channel within a preset frequency band along the frequency axis to form a measured energy arrival time series; calculate the difference between the measured energy arrival time series and the theoretical arrival time series to construct a spatial distortion vector; set a first threshold and a second threshold that are in a preset functional relationship with the reciprocal of the energy entropy; when the magnitude of the spatial distortion vector exceeds the first threshold and the number of components exceeding the second threshold reaches a preset number, determine that there is a defect in the corresponding pile section.

[0041] A preset frequency band, such as 40 to 60 kHz, is set based on the center frequency of the transmitted signal. In the defect energy indication diagram of each channel, the energy within this frequency band is accumulated along the frequency axis to obtain a one-dimensional curve of energy change over time. The time corresponding to the peak point of this curve is found and taken as the measured energy arrival time of that channel. This operation is repeated for all channels to form an N-dimensional measured energy arrival time series. Based on the known pile diameter, transducer placement, and standard concrete sound velocity, the theoretical sound wave propagation time of each channel is calculated, forming an N-dimensional theoretical arrival time series. The measured series is subtracted component by component from the theoretical series to obtain the spatial distortion vector V. First threshold Set as a constant Divide by energy entropy H, second threshold Set as a constant Divide by the energy entropy H, where and Let be an empirical constant. Calculate the Euclidean norm of the spatial distortion vector V, which is the square root of the sum of squares of all components. Statistically, vector V whose absolute value is greater than the second threshold is considered. The number of components. If the norm of vector V is greater than the first threshold. If the number of components exceeding the second threshold reaches a preset number, such as 3, then the pile section is determined to have a defect.

[0042] In an optional embodiment, calculating the difference between the measured energy arrival time series and the theoretical arrival time series to construct a spatial distortion vector includes: The theoretical arrival time series The i-th component is determined by the linear geometric distance between the ultrasonic transmitting and receiving probes corresponding to the i-th channel. and the preset sound velocity of the pile concrete Through formula Calculated; Measured energy arrival time series The components and theoretical arrival time series Subtracting the corresponding components yields the spatial distortion vector. .

[0043] Time delay is represented by comparing the actual sound wave propagation time with the ideal propagation time. A theoretical arrival time series is constructed. It is assumed that six ultrasonic testing channels are arranged on the pile cross-section, and the preset concrete sound velocity... The speed is 4000 meters per second. For channel 1, the probe spacing is... If the distance is 0.5m, then the theoretical arrival time is... =0.000125s, or 125μs. Calculate the theoretical arrival time for all 6 channels sequentially, forming a vector containing 6 elements, for example... =125,130,125,130,128,128μs.

[0044] The measured arrival time series of energy peak values ​​for each channel Compare with theoretical values. Assume the measured time series is obtained through signal processing. The values ​​were 127, 145, 126, 132, 150, and 130 μs. This was achieved by... Subtract each element from The elements at corresponding positions are used to construct the spatial distortion vector ΔT. The calculation result is: ΔT = 2, 15, 1, 2, 22, 2 μs. The above vector visually shows the acoustic delay of each channel. Delays are observed in channels 2 and 5, which may be caused by internal defects in the pile, such as... Figure 3 .

[0045] In an optional embodiment, setting the first threshold and the second threshold, which are preset functionally related to the reciprocal of the energy entropy, includes: Energy entropy is H, first threshold From the linear formula Calculate the second threshold. From the linear formula Calculation, where This is a constant obtained by calibrating standard defect-free pile experimental data.

[0046] Specifically, a set of defect judgment thresholds was established, which can adjust the strictness of the judgment according to the complexity of the signal. Four constants need to be calibrated through a large amount of experimental data. Assuming that the determination is achieved through testing and analysis of multiple standard piles... It is 20. It is 5. It is 8. The constant is 2. Once determined, this constant can be used for subsequent detection. When the energy entropy H is high, it indicates that the energy distribution is dispersed and the signal uncertainty is large. In this case, the threshold should be set lower, i.e., more stringent, to detect potential minor anomalies.

[0047] In a practical test, assume the energy entropy H calculated from the defect energy indicator map is 2.5. According to the formula, the first threshold... The calculation result is 13. Second threshold. The calculated result is 5.2. If the detection is performed at another location, the signal energy is highly concentrated, and the calculated energy entropy H is only 1.0, indicating high signal quality and strong reliability. In this case, the first threshold... The calculated result is 25; second threshold The calculation result is 10. For example... Figure 4 When the signal quality is good, the judgment criteria will be relaxed, allowing for greater time distortion, thereby reducing misjudgments.

[0048] In an optional embodiment, determining that the corresponding pile section has a defect when the magnitude of the spatial distortion vector exceeds a first threshold and the number of components exceeding a second threshold reaches a preset number includes: The square root of the sum of the squares of all components of the spatial distortion vector is taken as the magnitude of the spatial distortion vector. When the magnitude of the spatial distortion vector is greater than the first threshold, and the number of components in the vector whose absolute value is greater than the second threshold reaches the preset number of 3 or more, a defect is determined to exist.

[0049] The magnitude of the spatial distortion vector is calculated by combining two criteria: the overall magnitude of the distortion vector and the universality of local anomalies. Assume the spatial distortion vector ΔT obtained from a cross-section is 2,15,1,2,22,2μs. The norm, or modulus, is approximately 26.8. Assume the first threshold is calculated from the energy entropy. The value is 13, the second threshold. It is 5.2.

[0050] The calculation result is compared with two thresholds. The first criterion is to check whether the vector magnitude exceeds the first threshold. 26.8 > 13, so the first condition is met, indicating that the overall time distortion of this section is relatively large. The second criterion is that the absolute value in the statistical vector exceeds the second threshold. The number of components. In vector ΔT, the components with an absolute value greater than 5.2 are 15 and 22, a total of 2. The preset requirement is that the number must reach 3. Since there are 2, which is less than 3, the second condition is not met. Because both conditions must be met simultaneously to be considered a defect, although the overall distortion is large, the abnormal channels are not sufficiently dispersed, so it is determined that there is no prominent concentrated defect in this section. If the vector is 10, 15, 8, 2, 22, 9, then the magnitude exceeds And more than If there are 5 components, and 3 or more are greater than or equal to 3, then it will be judged as having a defect.

[0051] In the second embodiment, an online monitoring system for the quality of cast-in-place piles based on an ultrasonic array is also provided, comprising the following modules: The acquisition module is used to transmit and receive ultrasonic waves using an ultrasonic array arranged in the cast-in-place pile, and to acquire multi-channel acoustic signals. The selection module is used to perform generalized S-transform processing on each channel signal in the multi-channel acoustic signal to obtain their respective time-frequency matrices, and to calculate the cross-correlation coefficient between the channel signal and the preset defect-free reference signal. The calculation module is used to calculate the amplitude matrix and the instantaneous phase gradient matrix based on the time-frequency matrix; in the time-frequency domain, it fuses the neighborhood amplitude energy and neighborhood phase consistency factor of each point to generate a defect energy indication map of the pile cross section, wherein the weighting coefficient of the fusion is determined by the cross-correlation coefficient calculated above; and calculates the energy entropy of the defect energy indication map in the whole frequency band. The determination module is used to extract the peak energy times of each channel within a preset frequency band along the frequency axis in the defect energy indication diagram to form a measured energy arrival time series; calculate the difference between the measured energy arrival time series and the theoretical arrival time series to construct a spatial distortion vector; set a first threshold and a second threshold that are in a preset functional relationship with the reciprocal of the energy entropy; when the magnitude of the spatial distortion vector exceeds the first threshold and the number of components exceeding the second threshold reaches a preset number, it is determined that there is a defect in the corresponding pile section.

[0052] Those skilled in the art will understand that embodiments of the present invention can be provided as methods, systems, or computer program products. Therefore, the present invention can take the form of a completely hardware embodiment, a completely software embodiment, or an embodiment combining software and hardware aspects. Furthermore, the present invention can take the form of a computer program product embodied on one or more computer-usable storage media (including, but not limited to, disk storage, compact disc read-only memory (CD-ROM), optical storage, etc.) containing computer-usable program code.

[0053] This invention is described with reference to flowchart illustrations and / or block diagrams of methods, apparatus (systems), and computer program products according to embodiments of the invention. It will be understood that each block of the flowchart illustrations and / or block diagrams, and combinations of blocks in the flowchart illustrations and / or block diagrams, can be implemented by computer program instructions. These computer program instructions can be provided to a processor of a general-purpose computer, special-purpose computer, embedded processor, or other programmable data processing apparatus to produce a machine, such that the instructions, which execute via the processor of the computer or other programmable data processing apparatus, generate instructions for implementing the flowchart illustrations and / or block diagrams. Figure 1 One or more processes and / or boxes Figure 1 A device that provides the functions specified in one or more boxes.

[0054] The above description is merely an embodiment of this application and is not intended to limit the scope of this application. Various modifications and variations can be made to this application by those skilled in the art. Any modifications, equivalent substitutions, improvements, etc., made within the spirit and principles of this application should be included within the scope of the claims of this application.

Claims

1. A method for online monitoring of the quality of cast-in-place piles based on ultrasonic arrays, characterized in that, Includes the following steps: An ultrasonic array arranged inside the cast-in-place pile is used to transmit and receive ultrasonic waves to obtain multi-channel acoustic signals. The generalized S-transform is performed on each channel signal in the multi-channel acoustic signal to obtain their respective time-frequency matrices, and the cross-correlation coefficient between the channel signal and the preset defect-free reference signal is calculated. Based on the time-frequency matrix, the amplitude matrix and the instantaneous phase gradient matrix are calculated respectively; in the time-frequency domain, the neighborhood amplitude energy and neighborhood phase consistency factor of each point are fused to generate a defect energy indication map of the pile section, wherein the weighting coefficient of the fusion is determined by the cross-correlation coefficient calculated above; the energy entropy of the defect energy indication map in the whole frequency band is calculated. In the defect energy indication diagram, the peak energy times of each channel within a preset frequency band are extracted along the frequency axis to form a measured energy arrival time series; the difference between the measured energy arrival time series and the theoretical arrival time series is calculated to construct a spatial distortion vector; a first threshold and a second threshold are set that are in a preset functional relationship with the reciprocal of the energy entropy; when the magnitude of the spatial distortion vector exceeds the first threshold and the number of components exceeding the second threshold reaches a preset number, it is determined that there is a defect in the corresponding pile section.

2. The method according to claim 1, characterized in that, The generalized S-transform employs an asymmetric Gaussian window function. The skewness of the window function has a preset nonlinear mapping relationship with the local energy gradient of the signal within the time window. The frequency adjustment factor of the generalized S-transform is selected based on the initial signal-to-noise ratio of the channel signal, including: The skewness of the window function Through formula Determine, where G(t,f) is the local energy gradient of the signal within the time window. and The frequency adjustment factor p of the generalized S-transform is a preset constant; the frequency adjustment factor p of the generalized S-transform is expressed by the formula... Confirmed, among which Let be the initial signal-to-noise ratio of the channel signal, and k and b be preset constants.

3. The method according to claim 1, characterized in that, For each point (t,f) in the time-frequency matrix, select a 3×3 rectangular neighborhood around it; Calculate the sum of squared magnitudes of all 9 points in the neighborhood, which is taken as the neighborhood magnitude energy E(t,f) of the point; Calculate the magnitude of the vector sum of the instantaneous phase gradients of all 9 points in the neighborhood, and use it as the neighborhood phase consistency factor P(t,f) of the point.

4. The method according to claim 3, characterized in that, The weighting coefficients for the fusion are determined by the cross-correlation coefficients calculated above, including: Let the cross-correlation coefficient be R, then the weighting coefficient of the neighborhood magnitude energy... The weighting coefficient of the neighborhood phase consistency factor ; Normalize all calculated neighborhood magnitude energies E(t,f) and neighborhood phase coherence factors P(t,f) to obtain normalized neighborhood magnitude energies. Phase consistency factor with neighborhood ; The energy value D(t,f) of point (t,f) in the defect energy indicator map is obtained through a weighted fusion formula. Calculated.

5. The method according to claim 1, characterized in that, The calculation of the energy entropy of the defect energy indicator map across the entire frequency band includes: Defect energy indicator map The total energy is obtained by summing all the energy values. And calculate the probability of each time frequency point. ; The energy entropy is calculated using the Shannon entropy formula.

6. The method according to claim 1, characterized in that, The calculation of the difference between the measured energy arrival time series and the theoretical arrival time series, and the construction of the spatial distortion vector, includes: The theoretical arrival time series The i-th component is determined by the linear geometric distance between the ultrasonic transmitting and receiving probes corresponding to the i-th channel. and the preset sound velocity of the pile concrete Through formula Calculated; Measured energy arrival time series The components and theoretical arrival time series Subtracting the corresponding components yields the spatial distortion vector. .

7. The method according to claim 1, characterized in that, The first and second thresholds, which are preset to be functionally related to the reciprocal of the aforementioned energy entropy, include: Energy entropy is H, first threshold From the linear formula Calculate the second threshold. From the linear formula Calculation, where This is a constant obtained by calibrating standard defect-free pile experimental data.

8. The method according to claim 1, characterized in that, When the magnitude of the spatial distortion vector exceeds a first threshold, and the number of components exceeding a second threshold reaches a preset number, it is determined that the corresponding pile section has a defect, including: The square root of the sum of the squares of all components of the spatial distortion vector is taken as the magnitude of the spatial distortion vector. When the magnitude of the spatial distortion vector is greater than the first threshold, and the number of components in the vector whose absolute value is greater than the second threshold reaches the preset number of 3 or more, a defect is determined to exist.

9. An online monitoring system for the quality of cast-in-place piles based on an ultrasonic array, characterized in that, Includes the following modules: The acquisition module is used to transmit and receive ultrasonic waves using an ultrasonic array arranged in the cast-in-place pile, and to acquire multi-channel acoustic signals. The selection module is used to perform generalized S-transform processing on each channel signal in the multi-channel acoustic signal to obtain their respective time-frequency matrices, and to calculate the cross-correlation coefficient between the channel signal and the preset defect-free reference signal. The calculation module is used to calculate the amplitude matrix and the instantaneous phase gradient matrix based on the time-frequency matrix; in the time-frequency domain, it fuses the neighborhood amplitude energy and neighborhood phase consistency factor of each point to generate a defect energy indication map of the pile cross section, wherein the weighting coefficient of the fusion is determined by the cross-correlation coefficient calculated above; and calculates the energy entropy of the defect energy indication map in the whole frequency band. The determination module is used to extract the peak energy times of each channel within a preset frequency band along the frequency axis in the defect energy indication diagram to form a measured energy arrival time series; calculate the difference between the measured energy arrival time series and the theoretical arrival time series to construct a spatial distortion vector; set a first threshold and a second threshold that are in a preset functional relationship with the reciprocal of the energy entropy; when the magnitude of the spatial distortion vector exceeds the first threshold and the number of components exceeding the second threshold reaches a preset number, it is determined that there is a defect in the corresponding pile section.

10. The system according to claim 9, characterized in that, The generalized S-transform employs an asymmetric Gaussian window function. The skewness of the window function has a preset nonlinear mapping relationship with the local energy gradient of the signal within the time window. The frequency adjustment factor of the generalized S-transform is selected based on the initial signal-to-noise ratio of the channel signal, including: The skewness of the window function Through formula Determine, where G(t,f) is the local energy gradient of the signal within the time window. and The frequency adjustment factor p of the generalized S-transform is a preset constant; the frequency adjustment factor p of the generalized S-transform is expressed by the formula... Confirmed, among which Let be the initial signal-to-noise ratio of the channel signal, and k and b be preset constants.

Images