Pipeline wall thickness detection method and device

Abstract

The invention provides a pipeline wall thickness detection method and device, and the method comprises the steps: S1, enabling a robot to carry an electromagnetic ultrasonic pipeline detection cabin, entering the interior of a pipeline, obtaining an echo signal, and processing the noise in the echo signal; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness measuring probes; s2, calculating ultrasonic thickness measurement data of the pipeline according to the echo time interval of the echo signal; s3, judging whether missing values and / or abnormal values exist in the ultrasonic thickness measurement data or not; if yes, processing the ultrasonic thickness measurement data by adopting a nearest neighbor interpolation method; s4, constructing a detection surface of the pipeline, and determining a wall thickness detection surface detection value of the pipeline according to the size, corresponding to the thickness measurement data, on the circumference of each detection surface; the thickness measurement data is obtained through the thickness measurement probe. Therefore, the reliability of a measurement result is improved, and the method has positive significance in improving the safety and the stability of pipeline equipment.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of non-destructive testing, in particular to a pipeline wall thickness detection method and device. BACKGROUND

[0002] Electromagnetic acoustic technology is a widely used non-destructive testing technology for pipeline wall thickness detection. However, in the actual detection process, the problem of noise mixed in the echo signal is often encountered, including random noise, impulse noise and high frequency noise, etc. These noises will affect the accuracy and reliability of the measurement results. SUMMARY

[0003] The present application provides a pipeline wall thickness detection method and device for removing various noises in the echo signal and constructing the detection surface of the pipeline to determine the pipeline wall thickness, improving the accuracy and reliability of the measurement results.

[0004] In a first aspect, the present application provides a pipeline wall thickness detection method, the method comprising:

[0005] S1, an electromagnetic acoustic pipeline detection bin is carried by a robot to enter the inside of a pipeline to obtain an echo signal, and noises in the echo signal are processed; the electromagnetic acoustic pipeline detection bin is provided with a plurality of thickness measurement probes;

[0006] S2, ultrasonic thickness measurement data of the pipeline is calculated according to an echo time interval of the echo signal;

[0007] S3, it is judged whether the ultrasonic thickness measurement data appears missing values and / or abnormal values; if yes, the nearest neighbor interpolation method is used to process the ultrasonic thickness measurement data;

[0008] S4, a detection surface of the pipeline is constructed, and a wall thickness detection surface detection value of the pipeline is determined according to the size of the thickness measurement data corresponding to each detection surface circumference; the thickness measurement data is obtained by the thickness measurement probe.

[0009] Optionally, the step S4 comprises:

[0010] S41, the detection surface is constructed with the starting position of the pipeline as a detection starting point, a point on the center axis of the pipeline as a center, the inner diameter radius of the pipeline as a radius, and a detection step of the electromagnetic acoustic pipeline detection bin;

[0011] S42, the thickness measurement data corresponding to each detection surface circumference is obtained;

[0012] S43, the minimum value of all the thickness measurement data is taken as the wall thickness detection surface detection value.

[0013] Optionally, the noise comprises random noise, pulsating noise, medium frequency noise and high frequency noise; the S2 comprises:

[0014] S21, processing the random noise in the echo signal using an average filtering method to obtain a first filtered echo signal;

[0015] S22, processing the pulsating noise and the high frequency noise in the first filtered echo signal using a FIR filter to obtain a second filtered echo signal;

[0016] S23, suppressing the medium frequency noise and the high frequency noise similar to the echo signal in the second filtered echo signal by a wavelet transform method to obtain a noise-eliminated echo signal.

[0017] Optionally, the step S23 comprises:

[0018] S231, wavelet transforming the second filtered echo signal to obtain wavelet coefficients;

[0019] S232, using wavelet soft threshold noise reduction technology to reduce the noise coefficients in the wavelet coefficients to zero or close to zero to obtain noise-eliminated wavelet coefficients;

[0020] S233, inversely transforming the noise-eliminated wavelet coefficients into the noise-eliminated echo signal by wavelet inverse transform.

[0021] In a second aspect, the present application provides a pipeline wall thickness detection device, comprising:

[0022] An echo signal acquisition module is configured to enter the inside of a pipeline by a robot carrying an electromagnetic ultrasonic pipeline detection bin, acquire an echo signal, and process noise in the echo signal; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness measurement probes;

[0023] An ultrasonic thickness measurement data calculation module is configured to calculate ultrasonic thickness measurement data of the pipeline according to an echo time interval of the echo signal;

[0024] A data state judgment module is configured to judge whether the ultrasonic thickness measurement data has missing values and / or abnormal values; if so, the ultrasonic thickness measurement data is processed by a nearest neighbor interpolation method;

[0025] A detection module is configured to construct a detection surface of the pipeline, and determine a wall thickness detection surface detection value of the pipeline according to the size of the thickness measurement data corresponding to each detection surface circumference; the thickness measurement data is acquired by the thickness measurement probes.

[0026] Optionally, the detection module comprises:

[0027] A detection surface construction sub-module is configured to construct the detection surface with the starting position of the pipeline as a detection starting point, with a point on the pipeline center axis as a circle center, with the inner diameter radius of the pipeline as a radius, and with the electromagnetic ultrasonic pipeline detection bin detection step.

[0028] A thickness data acquisition sub-module is configured to acquire the corresponding thickness data on the circumference of each detection surface.

[0029] A detection sub-module is configured to take the minimum value of all the thickness data as the wall thickness detection surface detection value.

[0030] Optionally, the noise includes random noise, pulsating noise, medium frequency noise, and high frequency noise; and the ultrasonic thickness data calculation module includes:

[0031] A noise processing sub-module is configured to process the random noise in the echo signal using an average filtering method to obtain a first filtered echo signal.

[0032] A second filtering sub-module is configured to process the pulsating noise and the high frequency noise in the first filtered echo signal using a FIR filter to obtain a second filtered echo signal.

[0033] A suppression sub-module is configured to suppress the medium frequency noise and the high frequency noise similar to the echo signal in the second filtered echo signal by a wavelet transform method to obtain a noise-eliminated echo signal.

[0034] Optionally, the suppression sub-module includes:

[0035] A transform unit is configured to perform wavelet transform on the second filtered echo signal to obtain wavelet coefficients.

[0036] A noise reduction unit is configured to reduce noise coefficients in the wavelet coefficients to zero or close to zero by using a wavelet soft threshold noise reduction technology to obtain noise-eliminated wavelet coefficients.

[0037] An inverse transform unit is configured to perform inverse wavelet transform on the noise-eliminated wavelet coefficients to obtain the noise-eliminated echo signal.

[0038] In a third aspect, the present application provides an electronic device including a processor and a memory, wherein the memory stores computer readable instructions, and when the computer readable instructions are executed by the processor, the steps in the method provided in the first aspect are executed.

[0039] In a fourth aspect, the present application provides a storage medium having a computer program stored thereon, which, when executed by a processor, performs the steps of the method according to the first aspect.

[0040] From the above technical solutions, the present application has the following advantages:

[0041] The present application provides a pipeline wall thickness detection method and device, the method comprising: S1, an electromagnetic ultrasonic pipeline detection bin is carried by a robot to enter the inside of a pipeline to obtain an echo signal, and noise in the echo signal is processed; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness measurement probes; S2, ultrasonic thickness measurement data of the pipeline is calculated according to an echo time interval of the echo signal; S3, it is judged whether the ultrasonic thickness measurement data appears missing values and / or abnormal values; if so, the nearest neighbor interpolation method is used to process the ultrasonic thickness measurement data; S4, a detection surface of the pipeline is constructed, and a wall thickness detection surface detection value of the pipeline is determined according to the thickness measurement data corresponding to the size of each detection surface circumference; the thickness measurement data is obtained by the thickness measurement probe. By removing various noises in the echo signal and constructing the detection surface of the pipeline to determine the pipeline wall thickness, the reliability of the measurement result is improved, which has a positive significance for improving the safety and stability of the pipeline equipment. BRIEF DESCRIPTION OF DRAWINGS

[0042] In order to more clearly illustrate the technical solutions in the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiments or prior art description. Obviously, the drawings in the following description only constitute some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor.

[0043] Figure 1 The flow step diagram of the pipeline wall thickness detection method embodiment one of the present application;

[0044] Figure 2 The flow step diagram of the pipeline wall thickness detection method embodiment two of the present application;

[0045] Figure 3 The echo signal preprocessing method flow chart;

[0046] Figure 4 The self-adaptive threshold function curve chart;

[0047] Figure 5 The Gaussian echo model simulation curve with -5dB noise;

[0048] Figure 6 The signal preprocessing algorithm flow chart;

[0049] Figure 7 for filtering out random noise signals;

[0050] Figure 8 for a band-pass filtered signal graph;

[0051] Figure 9 for an adaptive threshold wavelet processed signal graph

[0052] Figure 10 for a schematic diagram of pipeline detection data processed by data processing;

[0053] Figure 11 for a schematic diagram of pipeline detection data processed by data processing;

[0054] Figure 12 for a schematic diagram of adjacent echo time intervals;

[0055] Figure 13 for a structural block diagram of an embodiment of a pipeline wall thickness detection device of the present application. DETAILED DESCRIPTION

[0056] The embodiment of the present application provides a pipeline wall thickness detection method and device, which is used for removing various noises in echo signals and constructing a detection surface of a pipeline to determine the pipeline wall thickness, and improves the accuracy and reliability of measurement results.

[0057] In order to make the purposes, features and advantages of the present application more obvious and easy to understand, the technical solutions in the embodiments of the present application will be described clearly and completely below with reference to the drawings in the embodiments of the present application. Obviously, the following described embodiments are only some of the embodiments of the present application, but not all the embodiments. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the protection scope of the present application.

[0058] Embodiment one, please refer to Figure 1 , Figure 1 for a flow step diagram of an embodiment one of a pipeline wall thickness detection method of the present application, comprising:

[0059] S1, an electromagnetic ultrasonic pipeline detection bin is carried by a robot to enter the inside of a pipeline to obtain echo signals, and noises in the echo signals are processed; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness measurement probes;

[0060] S2, ultrasonic thickness measurement data of the pipeline is calculated according to echo time intervals of the echo signals;

[0061] In an optional embodiment, the noises include random noises, pulsating noises, intermediate frequency noises and high frequency noises; the S2 includes:

[0062] S21, the random noise in the echo signal is processed using an average filtering method to obtain a first filtered echo signal;

[0063] S22, the pulsating noise and the high-frequency noise in the first filtered echo signal are processed using a FIR filter to obtain a second filtered echo signal;

[0064] S23, the medium-frequency noise and the high-frequency noise similar to the echo signal in the second filtered echo signal are suppressed by a wavelet transform method to obtain a noise-eliminated echo signal.

[0065] S3, it is judged whether the ultrasonic thickness data appears missing value and / or abnormal value; if yes, the nearest neighbor interpolation method is used to process the ultrasonic thickness data;

[0066] S4, a detection surface of the pipeline is constructed, and a wall thickness detection surface detection value of the pipeline is determined according to the size of the thickness data corresponding to each detection surface on the circumference; the thickness data is obtained by the thickness probe.

[0067] In the pipeline wall thickness detection method provided in the embodiment of the application, the steps include: S1, an electromagnetic ultrasonic pipeline detection bin is carried by a robot to enter the inside of a pipeline to obtain an echo signal, and the noise in the echo signal is processed; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness probes; S2, ultrasonic thickness data of the pipeline is calculated according to the echo time interval of the echo signal; S3, it is judged whether the ultrasonic thickness data appears missing value and / or abnormal value; if yes, the nearest neighbor interpolation method is used to process the ultrasonic thickness data; S4, a detection surface of the pipeline is constructed, and a wall thickness detection surface detection value of the pipeline is determined according to the size of the thickness data corresponding to each detection surface on the circumference; the thickness data is obtained by the thickness probe. By removing various noises in the echo signal and constructing the detection surface of the pipeline to determine the pipeline wall thickness, the reliability of the measurement result is improved, which has a positive significance for improving the safety and stability of the pipeline equipment.

[0068] Embodiment two, please refer to Figure 2 , Figure 2 is a flow step diagram of the pipeline wall thickness detection method embodiment two of the application, which includes:

[0069] Step S201, an electromagnetic ultrasonic pipeline detection bin is carried by a robot to enter the inside of a pipeline to obtain an echo signal, and the noise in the echo signal is processed; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness probes;

[0070] The pipe wall thickness detection method of the present invention is based on processing the effective signal frequency within the electromagnetic ultrasonic thickness measurement echo signal. The effective signal frequency refers to the frequency range of the signal after the ultrasonic wave is transmitted from the sensor, propagates through the object, and is reflected, returning to the sensor. For homogeneous materials such as metals, ultrasonic wave propagation speed is high, and the frequency is generally high, with the effective frequency range reaching several megahertz or even more than ten megahertz. Furthermore, the effective frequency gradually decreases with increasing measurement depth.

[0071] Step S202, using an average filtering method to process the random noise in the echo signal to obtain an echo signal after primary filtering;

[0072] It should be noted that the current electromagnetic ultrasonic transducer has a low transducer efficiency, resulting in a low signal-to-noise ratio of its echo signal. The effective signal is mixed with noise signals such as random noise, pulsating noise, and medium and high frequency noise, which results in discontinuous or incomplete signal sampling points, causing errors in positioning and data analysis, and thus affecting the accuracy and reliability of detection. There are many reasons for the generation of random noise in the electromagnetic ultrasonic thickness measurement echo signal, such as the noise of the equipment itself, the instability of the sensor, and the interference of the measurement environment. These random noises will affect the accuracy and reliability of the measurement results. Furthermore, since the frequency of the pulsating noise and high-frequency noise is separated from the frequency of the electromagnetic ultrasonic thickness measurement echo signal, the embodiment of the present invention first uses the average filtering method to eliminate part of the random noise, and then designs a digital bandpass filter (FIR filter) to filter out part of the low-frequency pulsating noise and high-frequency noise, and finally uses the wavelet transform method to decompose and reduce noise, improve the signal-to-noise ratio, and provide a basis for subsequent data analysis and processing. The processing process is as follows. Figure 3 The flow chart of the wave signal preprocessing method is shown in FIG.

[0073] Average filtering is a commonly used signal filtering method. Its basic idea is to average the signal to reduce the influence of random noise. The advantages of this method are simplicity, ease of implementation, good real-time performance, and it is suitable for processing random noise in signals.

[0074] In a specific implementation, the steps of filtering out random noise in the electromagnetic ultrasonic thickness measurement echo signal using the average filtering method include:

[0075] (1) Selecting the filter kernel size: Based on the characteristics of averaging filtering, a filter kernel of appropriate size should be selected. Generally, the appropriate kernel size is selected based on the actual measurement situation and signal characteristics. The larger the kernel, the more noise is attenuated, but the signal scale becomes larger.

[0076] (2) Average: for each sampling point, the arithmetic mean of the signal values in a certain size area centered on the point (including it) is calculated as the output value of the sampling point. Methods such as moving average, reverse average, cumulative average and exponential average can be used to average.

[0077] (3) Filtering: each output value obtained is taken as the new signal value of the corresponding sampling point, and the filtering of random noise is completed.

[0078] In the average filtering method, the filter kernel is a linear filter with a window size of n, and the mathematical expression is:

[0079]

[0080] Where y(i) represents the filtered signal, x(i) represents the original input signal, and n represents the filter kernel size, i.e. the window size. The formula indicates that the average value of the n signals in the window is taken as the filtered signal y(i), so as to achieve the purpose of removing noise.

[0081] In addition, in actual application, the smoothing degree and filtering effect can be adjusted by changing the filter kernel size to meet the specific signal processing requirements.

[0082] Step S203, using a FIR filter to process the pulsating noise and the high-frequency noise in the echo signal after the initial filtering, to obtain an echo signal after secondary filtering;

[0083] In specific implementation, the effective echo signal in the electromagnetic ultrasonic thickness measurement echo signal is concentrated in the medium-high frequency domain, so the frequency bands of the pulsating noise and the high-frequency noise have no intersection with the frequency band of the effective echo signal. Therefore, the medium-low frequency band and the high-frequency band noise signal of the pulsating noise and the high-frequency noise signal can be filtered out using a band-pass filter. Considering the accuracy and stability, a FIR filter which does not distort the phase after signal processing is selected.

[0084] The difference form of the FIR filter can be expressed as:

[0085]

[0086] Where x(i) is the original signal to be processed, h(k) is the FIR filter coefficient, y'(i) is the processed signal, I is the number of taps of the filter, and the order of the filter is I-1.

[0087] Step S204, wavelet transforming the echo signal after the secondary filtering to obtain wavelet coefficients;

[0088] Step S205, for the noise coefficient in the wavelet coefficient, the wavelet soft threshold denoising technology is used to reduce it to zero or close to zero, and the wavelet coefficient after eliminating noise is obtained.

[0089] Step S206, the wavelet coefficient after eliminating noise is inversely transformed into the echo signal after eliminating noise by wavelet inverse transform.

[0090] It should be noted that in electromagnetic ultrasonic measurement, the middle-high frequency noise similar to the echo signal is a common signal interference. If this kind of noise is not processed, it will affect the accuracy and reliability of the signal, and further reduce the performance of the whole measurement system. The wavelet transform can suppress the middle-high frequency noise similar to the echo signal in the electromagnetic ultrasonic thickness measurement echo signal, improve the quality and accuracy of the signal, and improve the performance of the measurement system.

[0091] Specifically, the formula of continuous wavelet transform is:

[0092]

[0093] Wherein, a is the scale, b is the translation coefficient, ψ is the wavelet base function, f(x) is the original signal, and W(a,b) is the wavelet coefficient.

[0094] The formula of discrete wavelet transform is:

[0095]

[0096] Wherein, X(j,n') is the low frequency coefficient with scale 2 j , Y(j,n') is the high frequency coefficient with scale 2 j , h(n') and g(n') are the impulse responses of low-pass and high-pass wavelet filters respectively.

[0097] The formula of wavelet inverse transform is:

[0098]

[0099] Wherein, J represents the decomposition level of wavelet transform, ψ (j,k) (x) is the wavelet base function.

[0100] The formula of wavelet threshold denoising is:

[0101]

[0102] Wherein, τ represents the threshold function, and different threshold processing methods can be used according to specific problems.

[0103] The specific steps of wavelet transform include:

[0104] (1) The electromagnetic ultrasonic thickness measurement echo signal is wavelet transformed to obtain wavelet coefficients.

[0105] (2) The wavelet coefficients are classified into noise coefficients and echo coefficients according to their sizes.

[0106] (3) For the noise coefficients, a wavelet soft threshold denoising technology is used to reduce them to zero or close to zero to achieve the purpose of eliminating noise. For the echo coefficients, their original values are retained.

[0107] (4) The wavelet coefficients after noise elimination are inversely transformed into the electromagnetic ultrasonic thickness measurement echo signal after noise elimination through wavelet inverse transformation.

[0108] It should be noted that in the wavelet transformation, the selection of different wavelet basis functions, decomposition layers and threshold values will affect the noise elimination effect. Therefore, in actual application, appropriate wavelet basis functions and threshold values need to be selected according to the characteristics of the signal and application requirements to obtain better denoising effect.

[0109] For the medium and high frequency noise in the electromagnetic ultrasonic thickness measurement echo signal, a lower decomposition layer number can be selected, and the specific decomposition layer number needs to be selected in combination with the characteristics of the signal itself, processing purposes and data processing capacity and other factors. The common decomposition layer number is 4-6 layers, and the embodiment of the present application selects the decomposition layer number as 4 layers according to the medium and high frequency noise in the electromagnetic ultrasonic thickness measurement echo signal.

[0110] In addition to the above wavelet basis function selection affecting the processing effect, the threshold value selection algorithm also affects the wavelet denoising effect. Selecting an appropriate threshold value for denoising is also a key point of the wavelet threshold value method for removing medium and high frequency noise. The commonly used threshold functions are hard threshold function, soft threshold function and adaptive threshold function. The hard threshold function is a relatively rough wavelet denoising method, which is prone to artifacts. The soft threshold function has a constant error when reconstructing the signal. Based on this, the embodiment of the present application constructs an adaptive improved exponential threshold function method as an adaptive threshold function of the wavelet signal denoising method for medium and high frequency noise in the electromagnetic ultrasonic thickness measurement echo signal.

[0111] The mathematical formula of the adaptive improved exponential threshold function is:

[0112]

[0113] wherein W (j,k) represents the jth layer kth wavelet coefficient in the wavelet transformation, and λ is the threshold value. When the wavelet coefficient W (j,k)) is less than the threshold value λ, the wavelet coefficient is set to 0, indicating that the coefficient is discarded. When the wavelet coefficient W (j,k) is greater than or equal to the threshold value λ, the coefficient is retained. The adaptive threshold function curve (λ=3) newly constructed by the embodiment of the present application is as follows:Figure 4 The adaptive threshold function curve is shown in the figure.

[0114] Meanwhile, the exponential function is adopted in the embodiment of the application to replace the traditional smoothing function, so that the high frequency information and local features of the signal can be better preserved. Specifically, the Gaussian echo signal with -5dB noise is adopted as the preprocessed signal, and the performance indicators of the two methods are compared. The signal-to-noise ratio of the smoothing function is -4.85dB, the mean square error is 1.52e+00, and the peak signal-to-noise ratio is -0.24dB. The signal-to-noise ratio of the adaptive improved exponential threshold function is 4.79dB, the mean square error is 1.66e-01, and the peak signal-to-noise ratio is 7.81dB.

[0115] In terms of numerical values, the adaptive improved exponential threshold function method has a higher signal-to-noise ratio, a lower mean square error and a higher peak signal-to-noise ratio, which means that the effect of denoising by using the adaptive improved exponential threshold function is better.

[0116] It should be noted that the signal-to-noise ratio is the ratio of the signal intensity to the noise intensity, and a high signal-to-noise ratio value means that the signal intensity is higher than the noise intensity, and the signal quality is better. The signal-to-noise ratio of the smoothing function is -4.85dB, and the signal-to-noise ratio of the adaptive improved exponential threshold function is 4.79dB, which shows that the adaptive improved exponential threshold function is better in terms of noise elimination and signal preservation.

[0117] The mean square error is an indicator for measuring the difference between the original signal and the denoised signal in the denoising process, and the lower the mean square error, the smaller the difference, that is, the better the denoising effect. The mean square error of the smoothing function in the embodiment of the application is 1.52e+00, and the mean square error of the adaptive improved exponential threshold function is 1.66e-01. It can be seen that the adaptive improved exponential threshold function is more effective in reducing errors.

[0118] The peak signal-to-noise ratio is a commonly used indicator for evaluating the quality of an image or sound signal, and a high peak signal-to-noise ratio value usually means better denoising effect. In the embodiment of the application, the peak signal-to-noise ratio of the smoothing function is -0.24dB, and the peak signal-to-noise ratio of the adaptive improved exponential threshold function is 7.81dB, which shows that the adaptive improved exponential threshold function is more excellent in signal quality recovery.

[0119] In summary, since the adaptive improved exponential threshold function is superior to the smoothing function in terms of signal-to-noise ratio, mean square error and peak signal-to-noise ratio, the denoising effect is better by using the adaptive improved exponential threshold function in this specific problem.

[0120] In addition, the Gaussian echo model is selected in the embodiment of the application to simulate the electromagnetic ultrasonic echo signal, and the generated signal is as shown in Figure 5The Gaussian echo model simulation curve with -5dB noise is shown. According to the set parameters and the Gaussian echo model expression of the electromagnetic ultrasonic echo signal, a signal with a bandwidth factor of 25 (MHz) 2 , a center frequency of 5MHz, an arrival time of 1 microsecond, a set interval time of 1000ns, and -5dB signal-to-noise ratio noise added to the signal is generated.

[0121] The Gaussian echo model expression of the electromagnetic ultrasonic echo signal is:

[0122]

[0123] Where s(t) represents the echo signal, a represents the amplitude of the signal, t0 represents the center time of the signal, σ represents the standard deviation of the signal, n(t) represents the noise signal, and f c represents the carrier frequency of the signal.

[0124] In a specific implementation, for the three kinds of noise in the simulated electromagnetic ultrasonic thickness measurement echo signal, different denoising schemes are designed according to the characteristics of the noise. In order to improve the operation efficiency and reduce the subsequent data processing amount, the random noise is suppressed first, then the high and low frequency noise is suppressed, and finally the wavelet denoising is used to suppress the medium and high frequency noise. The signal preprocessing algorithm is as shown in the signal preprocessing algorithm flowchart, which includes: Figure 6

[0125] (1) The 6-time average filtering method is used to filter out the random noise in the electromagnetic ultrasonic thickness measurement echo signal, as shown in the random noise signal filtering diagram. Figure 7

[0126] (2) The 4th order FIR band-pass filter is used to filter the signal with noise, as shown in the band-pass filtered signal diagram. Figure 8

[0127] (3) The wavelet transform is used for medium and high frequency noise reduction, the wavelet base function is selected as 6db, the wavelet decomposition is 4 layers, and the wavelet threshold processing function is the adaptive threshold processing function, as shown in the adaptive threshold wavelet processed signal diagram. Figure 9

[0128] After the above process, under the condition of a signal-to-noise ratio of -5dB, the random noise signal-to-noise ratio is improved from -5dB to 2.6dB; the low frequency pulsating noise and high frequency white noise outside 3-6MHz are filtered out; finally, the four-layer decomposed wavelet function is used, the improved wavelet threshold function is used for the final preprocessing and noise reduction of the signal, the medium and high frequency noise similar to the effective signal frequency is suppressed, and the root mean square error is finally obtained. 0.05. After the preprocessing of the original signal, the signal-to-noise ratio of the original signal is greatly improved, which effectively creates conditions for subsequent processing. ​​​​

[0129] Step S207, judging whether the ultrasonic thickness measurement data has missing values and / or abnormal values; if yes, adopting the nearest neighbor interpolation method to process the ultrasonic thickness measurement data;

[0130] Step S208, taking the starting position of the pipeline as a detection starting point, taking a point on the center axis of the pipeline as a circle center, taking the inner diameter radius of the pipeline as a radius, and taking the electromagnetic ultrasonic pipeline detection bin detection step to construct the detection surface;

[0131] Step S209, obtaining the corresponding thickness measurement data on the circumference of each detection surface;

[0132] Step S210, taking the minimum value of all the thickness measurement data as the wall thickness detection surface detection value.

[0133] In the embodiment of the present application, assuming that a probe is arranged on the electromagnetic ultrasonic detection bin carried by the robot, the detection point value of each detection probe of the detection bin is P y (y = 1, …, a), after the pipeline internal detection robot carrying the electromagnetic ultrasonic detection bin enters the pipeline, taking the starting position of the pipeline as a detection starting point, the robot carrying the detection bin moves along the pipeline axial direction in the pipeline, and a detection surface is formed by taking a point C y on the center axis of the pipeline as a circle center and taking the outer diameter D of the pipeline as a radius, the detection value (P1, …, P y , …, P a ) of each detection surface is obtained, and a detection surface is determined according to the detection step t of the electromagnetic ultrasonic detection probe, so that the whole pipeline has L / t detection surfaces, for each detection surface, a total of a pipeline wall thickness value is generated, and L / t detection surfaces have:

[0134] P j1 , …, P jy , …, P ja , j = 1, …, L / t.

[0135] Subsequently, data cleaning is performed on the obtained pipeline wall thickness values.

[0136] It should be noted that data cleaning is an indispensable link in the process of data analysis, and the quality of the result is directly related to the detection effect and the final conclusion. In actual operation, data cleaning usually accounts for 50% to 80% of the time of the analysis process. In the data processing process, cleaning work such as whether the data set has repetition, whether there is missing, whether the data has integrity and consistency, and whether there are outliers in the data is generally needed. When the above possible problems are found in the data, they need to be processed accordingly. The embodiment of the present application focuses on identifying and processing missing values and outliers in the pipe wall thickness value. The missing value refers to the missing indicator value of some observations in the data set. The existence of the missing value will also affect the result of data analysis and mining.

[0137] Generally speaking, when encountering missing values, replacement and interpolation methods can be used. The replacement method is to fill in the missing position with a certain statistical quantity. For example, for continuous variables, the mean or median can be used for replacement, and for discrete variables, the mode can be used for replacement. The interpolation method refers to predicting the missing value according to other non-missing variables or observations. Common interpolation methods include regression interpolation, K-nearest neighbor interpolation, and Lagrange interpolation.

[0138] The interpolation method generally follows the following principles:

[0139] (1) Nearest neighbor interpolation. Compare the existing data items of the data, and fill in the data of the most similar item;

[0140] (2) Regression method. Guess the missing value through the correlation between data variables by regression analysis;

[0141] (3) For time series problems, spline interpolation and Newton interpolation can be used.

[0142] The full name of the outlier in statistics is suspected outlier, also known as outlier. The analysis of the outlier is also called outlier analysis. The electromagnetic ultrasonic detection value outlier refers to the "extreme value" that appears in the detection process, and the data value looks abnormally large or small, and its distribution deviates significantly from the rest of the observations. Electromagnetic ultrasonic detection value outlier analysis is to test whether there are unreasonable data in the data. In data analysis, the existence of outliers cannot be ignored, and outliers cannot be simply excluded from data analysis.

[0143] For the detection of outliers, the box plot discrimination method can be used, that is, the method of using the inter-quartile range of the box plot to detect outliers. The inter-quartile range (IQR) is the difference between the upper and lower quartiles. The formula for calculating IQR is:

[0144] IQR = Q U -QL ;

[0145] Q U is the upper quartile, i.e. one quarter of the data is greater than Q L is the lower quartile, i.e. one quarter of the data is less than Q L -1.5xIQR or greater than Q U +1.5xIQR are considered outliers. The formula for detecting outliers is:

[0146]

[0147] The box plot provides a standard for identifying outliers. However, the method for handling outliers depends on the specific situation. Sometimes, outliers can be normal values, so in many cases, the possible reasons for outliers should be analyzed before determining how to handle outliers.

[0148] In general, outliers can be handled by interpolation, treating outliers as missing values and using the method for handling missing values. The advantage is that the existing data is used to replace or interpolate outliers.

[0149] In order to better understand the electromagnetic ultrasonic thickness measurement data acquisition and cleaning process of the embodiments of the present application, the following application is used as an example:

[0150] (1) Data acquisition: the total length of the pipeline is 10 m, the inner diameter of the pipeline is 300 mm, the design wall thickness of the pipeline is 10 mm, four electromagnetic ultrasonic thickness probes are arranged on the circumference of the electromagnetic ultrasonic detection bin entering the pipeline, the starting position of the pipeline is taken as the starting point of detection after the electromagnetic ultrasonic detection bin enters the pipeline, the detection bin moves forward in the pipeline along the pipeline axis, a detection surface is formed with the point on the center axis of the pipeline as the center and the inner diameter radius of the pipeline 150 mm as the radius, the detection step of the electromagnetic ultrasonic detection bin is 1 mm, one detection surface is determined for each detection step, and each detection surface corresponds to four detection values of the thickness probe.

[0151] (2) Data cleaning: The pipe wall thickness data detected by the four detection probes in the electromagnetic ultrasonic detection chamber are mostly between 7mm and 10mm. However, the detection process is subject to interference from the external environment, the detection chamber moves too fast, there is oil and dirt inside the pipe that causes the probe to lift off too much, and data is damaged or lost during data collection, storage and transmission, resulting in missing values ​​and abnormal values ​​in the detection data. Therefore, in the data processing process, the nearest neighbor interpolation method is used to deal with missing values ​​in the electromagnetic ultrasonic thickness measurement data; the box plot discrimination method is used to determine whether the electromagnetic ultrasonic thickness measurement data is an abnormal value. If it is an abnormal value, the nearest neighbor interpolation method is used to replace the abnormal value.

[0152] (3) Use MATLAB to write a method for processing missing values ​​and outliers to ensure the quality of the measurement data. Specifically:

[0153]

[0154]

[0155] (4) Data processing results: Pipeline inspection data before processing are as follows Figure 10 The data processing pipeline detection data diagram is shown in the figure. The detection data after processing is as follows Figure 11 The results show that this method can effectively identify and process missing values ​​and abnormal values ​​in electromagnetic ultrasonic detection values, thus ensuring the quality of detection data.

[0156] In addition, the basic principle for measuring pipe wall thickness in the embodiments of the present invention is the pulse reflection method. Thickness calculation primarily involves extracting the time interval Δt between adjacent echoes. Methods for extracting this time interval Δt include the gate method, the peak method, and the envelope method. The gate method sets an amplitude gate on the thickness measurement echo signal. When the amplitudes of two adjacent echoes first reach the preset gate, the time at which each occurs is recorded. The difference between these time values ​​is Δt.

[0157] The gate method is a commonly used method, and its specific process is as follows:

[0158] (1) Setting the gate: Set the appropriate gate width and height according to the material and thickness of the object being measured. The gate width should be as large as possible compared to the thickness of the object being measured, and the gate height should be fixed on the periodic waveform.

[0159] (2) Acquiring signals: Place the electromagnetic ultrasonic probe on the surface of the object to be measured to obtain the echo signal.

[0160] (3) Gate processing: Use the gate to process the signal and only retain the signal within the gate range.

[0161] (4) Peak detection: Perform peak detection on the processed signal to obtain the peak position and amplitude of the signal.

[0162] (5) Thickness calculation: Based on the material of the object being measured and the propagation speed of the electromagnetic ultrasonic wave, the thickness of the object being measured is calculated by calculating the signal propagation time and propagation distance.

[0163] A calculation program was written using MATLAB. The electromagnetic ultrasonic echo signal was obtained after filtering out random noise, pulsating noise, high-frequency noise, and medium- and low-frequency noise mentioned above. The thickness of the object being measured was calculated using the gate method. The main procedures are as follows:

[0164] (1) Set the gate width, height and threshold, specifically:

[0165] gate_width=30

[0166] gate_height=0.6

[0167] threshold=0.5

[0168] (2) Perform gate processing on the signal, specifically:

[0169] gate_start=60

[0170] gate_end=gate_start+gate_width

[0171] gated_data=data(gate_start:gate_end,:)

[0172] (3) Perform peak detection on the gated signal, specifically:

[0173] peak_value=max(abs(gated_data))

[0174] peak_position=find(abs(gated_data)==peak_value)

[0175] (4) Calculate the signal propagation time and distance, specifically:

[0176] probe_frequency=5e6;% probe operating frequency

[0177] signal_velocity=1.54e6;% electromagnetic ultrasonic wave propagation speed

[0178] time_delay=peak_position / probe_frequency;

[0179] distance = signal_velocity * time_delay / 2; % actual propagation distance is round trip, so divide by 2

[0180] (5) Calculate the thickness of the measured object, specifically:

[0181] thickness = distance / cosd(45); % assume signal propagation path is 45 degrees from normal to measured object

[0182] (6) Display the calculation results, specifically:

[0183] disp(['Measured object thickness is:'num2str(thickness)'mm']);

[0184] The program reads the electromagnetic ultrasonic thickness echo signal data, sets the parameters of the gate (width, height and threshold), and performs gate processing on the signal, only retaining the signal within the gate range. Then, the processed signal is peak detected to obtain the peak position and amplitude of the signal, and the signal propagation time and distance are calculated to calculate the thickness of the measured object. Finally, the calculation results are displayed. After running the MATLAB written calculation program, the time interval Δt between adjacent echoes is obtained, see Figure 12 the adjacent echo time interval diagram.

[0185] According to the material properties of the detected workpiece, the inherent sound speed W v of the measured workpiece material is obtained, the time difference Δt between adjacent echoes is obtained from the above calculation program, and the thickness d of the measured workpiece is calculated according to the following workpiece thickness calculation formula. Calculation formula:

[0186]

[0187] The method for detecting the wall thickness of a pipeline provided in the embodiment of the application comprises the following steps: S1, an electromagnetic ultrasonic pipeline detection bin is carried by a robot to enter the inside of a pipeline to obtain echo signals and process noise in the echo signals; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness measurement probes; S2, ultrasonic thickness measurement data of the pipeline is calculated according to echo time intervals of the echo signals; S3, it is judged whether the ultrasonic thickness measurement data has missing values and / or abnormal values; if yes, the ultrasonic thickness measurement data is processed by using a nearest neighbor interpolation method; S4, a detection surface of the pipeline is constructed, and a wall thickness detection surface detection value of the pipeline is determined according to the size of the thickness measurement data corresponding to each detection surface circumference; the thickness measurement data is obtained by the thickness measurement probes. The reliability of the measurement result is improved by removing various noise in the echo signals and constructing the detection surface of the pipeline to determine the wall thickness of the pipeline, which has a positive significance for improving the safety and stability of the pipeline equipment.

[0188] Embodiment three, please refer to Figure 13 , Figure 13 is a structural block diagram of the pipeline wall thickness detection device embodiment of the application, comprising:

[0189] The echo signal acquisition module 301 is configured to carry an electromagnetic ultrasonic pipeline detection bin by a robot to enter the inside of a pipeline to obtain echo signals and process noise in the echo signals; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness measurement probes;

[0190] The ultrasonic thickness measurement data calculation module 302 is configured to calculate ultrasonic thickness measurement data of the pipeline according to echo time intervals of the echo signals;

[0191] The data state judgment module 303 is configured to judge whether the ultrasonic thickness measurement data has missing values and / or abnormal values; if yes, the ultrasonic thickness measurement data is processed by using a nearest neighbor interpolation method;

[0192] The detection module 304 is configured to construct a detection surface of the pipeline, and determine a wall thickness detection surface detection value of the pipeline according to the size of the thickness measurement data corresponding to each detection surface circumference; the thickness measurement data is obtained by the thickness measurement probes.

[0193] In an optional embodiment, the detection module 304 comprises:

[0194] The detection surface construction submodule is configured to respectively take a starting position of the pipeline as a detection starting point, take a point on the center axis of the pipeline as a circle center, take an inner diameter radius of the pipeline as a radius, and construct the detection surface by using a detection step of the electromagnetic ultrasonic pipeline detection bin;

[0195] The thickness measurement data acquisition submodule is configured to obtain the thickness measurement data corresponding to each detection surface circumference.

[0196] The detection sub-module is configured to take the minimum value of all the thickness data as the wall thickness detection surface detection value.

[0197] In an optional embodiment, the noise includes random noise, pulsating noise, medium frequency noise and high frequency noise; and the ultrasonic thickness data calculation module 302 includes:

[0198] The noise processing sub-module is configured to process the random noise in the echo signal using an average filtering method to obtain a first filtered echo signal.

[0199] The secondary filtering sub-module is configured to process the pulsating noise and the high frequency noise in the first filtered echo signal using a FIR filter to obtain a second filtered echo signal.

[0200] The suppression sub-module is configured to suppress the medium frequency noise and the high frequency noise similar to the echo signal in the second filtered echo signal by a wavelet transform method to obtain a noise-eliminated echo signal.

[0201] In an optional embodiment, the suppression sub-module includes:

[0202] The transformation unit is configured to perform wavelet transform on the second filtered echo signal to obtain wavelet coefficients.

[0203] The noise reduction unit is configured to reduce noise coefficients in the wavelet coefficients to zero or close to zero by using a wavelet soft threshold noise reduction technology to obtain noise-eliminated wavelet coefficients.

[0204] The inverse transformation unit is configured to inversely transform the noise-eliminated wavelet coefficients into the noise-eliminated echo signal by wavelet inverse transform.

[0205] In embodiment four, the embodiments of the present application further provide an electronic device including a memory and a processor, the memory stores a computer program, and the computer program is executed by the processor to make the processor execute the steps of the pipe wall thickness detection method of any one of the embodiments.

[0206] In embodiment five, the embodiments of the present application further provide a computer storage medium storing a computer program, and the computer program is executed by the processor to implement the steps of the pipe wall thickness detection method of any one of the embodiments.

[0207] Those skilled in the art can clearly understand that, for the convenience and brevity of description, the specific working processes of the above-described system, device and unit can refer to the corresponding processes in the foregoing method embodiments, which will not be described here.

[0208] In several embodiments provided in the present application, it should be understood that the disclosed methods, devices, electronic devices and storage media can be implemented in other ways. For example, the above-described device embodiments are only illustrative, and for example, the division of the units is only a logical function division, and actual implementation can have another division manner, for example, a plurality of units or components can be combined or integrated into another system, or some features can be ignored or not executed. In addition, the shown or discussed mutual coupling or direct coupling or communication connection can be indirect coupling or communication connection through some interfaces, devices or units, which can be electrical, mechanical or other forms.

[0209] The units described as separate components can or can not be physically separated, and the components shown as units can or can not be physical units, i.e., they can be located in one place or distributed on multiple network units. Part or all of the units can be selected according to actual needs to achieve the purpose of the present embodiment.

[0210] In addition, the functional units in each embodiment of the present application can be integrated in one processing unit, or each unit can exist physically, or two or more units can be integrated in one unit. The integrated unit can be realized in the form of hardware or in the form of a software functional unit.

[0211] The integrated unit, if realized in the form of a software functional unit and sold or used as an independent product, can be stored in a computer-readable storage medium. Based on this understanding, the technical solutions of the present application essentially or the part that contributes to the prior art or the whole or part of the technical solutions can be embodied in the form of a software product. The computer software product is stored in a readable storage medium, including a plurality of instructions for causing a computer device (which can be a personal computer, a server, or a network device, etc.) to execute all or part of the steps of the method described in the various embodiments of the present application. The aforementioned readable storage medium includes: a U disk, a mobile hard disk, a read-only memory (ROM, Read-Only Memory), a random access memory (RAM, Random Access Memory), a magnetic disk or an optical disk, and various program code storage media.

[0212] The above-described embodiments are only used to illustrate the technical solutions of the present application, and are not intended to limit the present application; although the present application has been described in detail with reference to the foregoing embodiments, it should be understood by those skilled in the art that the technical solutions recorded in the foregoing embodiments can still be modified, or some technical features can be replaced by equivalent replacements; and these modifications or replacements do not make the essence of the corresponding technical solutions deviate from the spirit and scope of the technical solutions of the embodiments of the present application.

Claims

1. A method of detecting pipe wall thickness, the method comprising: The method comprises the following steps: ​ S1, the robot carries an electromagnetic ultrasonic pipeline detection bin, enters the inside of the pipeline to obtain echo signals, and processes noise in the echo signals; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness measurement probes; S2, according to the echo time interval of the echo signal, the ultrasonic thickness measurement data of the pipeline is calculated; S3, whether the ultrasonic thickness measurement data appears missing value and / or abnormal value is judged; if yes, the nearest neighbor interpolation method is used to process the ultrasonic thickness measurement data; S4, the detection surface of the pipeline is constructed, and the wall thickness detection surface detection value of the pipeline is determined according to the size of the thickness measurement data corresponding to each detection surface circumference; the thickness measurement data is obtained through the thickness measurement probe.

2. The method of claim 1, wherein The step S4 comprises: S41, respectively taking the starting position of the pipeline as a detection starting point, taking a point on the center axis of the pipeline as a center, taking the inner diameter radius of the pipeline as a radius, and taking the electromagnetic ultrasonic pipeline detection bin detection step to construct the detection surface; S42, the thickness measurement data corresponding to each detection surface circumference is obtained; S43, the minimum value of all the thickness measurement data is taken as the wall thickness detection surface detection value.

3. The method of claim 1, wherein The noise comprises random noise, pulsating noise, intermediate frequency noise and high frequency noise; the S2 comprises: S21, the random noise in the echo signal is processed by using the average filtering method to obtain the echo signal after primary filtering; S22, the pulsating noise and the high frequency noise in the echo signal after primary filtering are processed by using the FIR filter to obtain the echo signal after secondary filtering; S23, the intermediate frequency noise and the high frequency noise similar to the echo signal in the echo signal after secondary filtering are suppressed by the method of wavelet transform to obtain the echo signal after noise elimination.

4. The method of claim 3, wherein The step S23 comprises: S231, the echo signal after secondary filtering is subjected to wavelet transform to obtain wavelet coefficients; S232, the noise coefficients in the wavelet coefficients are reduced to zero or close to zero by using the wavelet soft threshold noise reduction technology to obtain the wavelet coefficients after noise elimination; S233, the wavelet coefficients after noise elimination are inversely transformed into the echo signal after noise elimination by wavelet inverse transform.

5. A device for detecting the wall thickness of a pipe, characterized in that The method comprises the following steps: An echo signal acquisition module is used to carry an electromagnetic ultrasonic pipeline detection bin by a robot, enter the inside of the pipeline to obtain echo signals, and process noise in the echo signals; the electromagnetic ultrasonic pipeline detection bin is provided with a plurality of thickness measurement probes; An ultrasonic thickness measurement data calculation module is used to calculate the ultrasonic thickness measurement data of the pipeline according to the echo time interval of the echo signal; A data state judgment module is used to judge whether the ultrasonic thickness measurement data appears missing value and / or abnormal value; if yes, the nearest neighbor interpolation method is used to process the ultrasonic thickness measurement data; A detection module is used to construct the detection surface of the pipeline, and determine the wall thickness detection surface detection value of the pipeline according to the size of the thickness measurement data corresponding to each detection surface circumference; the thickness measurement data is obtained through the thickness measurement probe.

6. The apparatus for detecting a pipe wall thickness according to claim 5, wherein The detection module comprises: The detection surface construction submodule is configured to construct the detection surface by taking the starting position of the pipeline as a detection starting point, taking a point on the pipeline center axis as a circle center, taking the inner diameter radius of the pipeline as a radius, and taking the electromagnetic ultrasonic pipeline detection bin detection step. The thickness data acquisition submodule is configured to acquire the corresponding thickness data on the circumference of each detection surface. The detection submodule is configured to take the minimum value of all the thickness data as the wall thickness detection surface detection value.

7. The apparatus for detecting a pipe wall thickness according to claim 5, wherein The noise includes random noise, pulsating noise, medium frequency noise, and high frequency noise. The noise processing submodule is configured to use an average filtering method to process the random noise in the echo signal to obtain a first filtered echo signal. The secondary filtering submodule is configured to use a FIR filter to process the pulsating noise and the high frequency noise in the first filtered echo signal to obtain a second filtered echo signal. The suppression submodule is configured to use a wavelet transform method to suppress the medium frequency noise and the high frequency noise similar to the echo signal in the second filtered echo signal to obtain a noise-eliminated echo signal.

8. The apparatus for detecting a pipe wall thickness according to claim 7, wherein The suppression submodule includes: The transformation unit is configured to perform wavelet transform on the second filtered echo signal to obtain wavelet coefficients. The noise reduction unit is configured to use a wavelet soft threshold noise reduction technology to reduce noise coefficients in the wavelet coefficients to zero or close to zero to obtain noise-eliminated wavelet coefficients. The inverse transformation unit is configured to perform inverse wavelet transform on the noise-eliminated wavelet coefficients to obtain the noise-eliminated echo signal.

9. An electronic device, comprising: The computer program is executed by the processor to run the method of any one of claims 1-4.

10. A storage medium having stored thereon a computer program, characterized in that The computer program is executed by the processor to run the method of any one of claims 1-4.

Images