Nondestructive ultrasonic flaw detection method and system for seamless oil casing

Abstract

The invention belongs to the technical field of ultrasonic detection, and discloses a nondestructive ultrasonic flaw detection method and system for a seamless oil casing. Comprising the steps of collecting sleeve feature data; the characteristic data of the casing pipe are analyzed, and ultrasonic parameters are dynamically adjusted; collecting ultrasonic flaw detection data; processing the ultrasonic flaw detection data to obtain ultrasonic processing data; performing feature extraction on the ultrasonic processing data, and extracting ultrasonic feature data; analyzing the ultrasonic characteristic data, and identifying the defects of the sleeve; according to the method, the defect problem in the oil casing can be effectively identified, the risk of misjudgment and missed judgment is reduced, the automation degree, precision and efficiency of nondestructive defect detection are remarkably improved, and a reliable guarantee is provided for full-life-cycle management of the seamless oil casing.

Description

Technical Field

[0001] The present invention relates to the field of ultrasonic testing technology, and more specifically, to a non-destructive ultrasonic flaw detection method and system for seamless oil casing. Background Art

[0002] Seamless oil casing is a key pipe in the oil and gas industry and is widely used in drilling and transportation processes. Due to its operation under high pressure, high temperature and corrosive environment, the quality and reliability of seamless oil casing directly affect the safety and economic benefits of oil and gas extraction. However, during production, transportation and use, seamless oil casing may have defects such as cracks, pores, inclusions, etc. If these defects are not detected and handled in time, they may lead to pipe failure and even cause serious safety accidents. Traditional non-destructive testing methods, such as magnetic particle testing and radiographic testing, can detect defects to a certain extent, but there are problems such as limited detection accuracy, low efficiency and high requirements for the detection environment. Ultrasonic flaw detection has become an important means of quality inspection for seamless oil casing due to its advantages such as high precision, rapidity and strong applicability.

[0003] The patent with publication number CN118817859A discloses a non-destructive ultrasonic flaw detection method and system for seamless oil casing; including: cleaning pretreatment, coupling agent application, automatic comprehensive scanning, signal reception and preprocessing, intelligent defect analysis, secondary verification of key areas, and result determination and recording; simplifies the non-destructive ultrasonic flaw detection process of seamless oil casing, including cleaning pretreatment, coupling agent application, automatic comprehensive scanning, signal reception and preprocessing, automatic analysis of ultrasonic signals after preprocessing using intelligent algorithms, key scanning and verification of suspected defective areas, and result determination and recording. By accurately setting detection parameters and intelligent analysis, efficient and accurate defect detection is achieved, and detection reports are automatically generated, thereby improving detection efficiency and reliability.

[0004] However, in the above-mentioned technology, the inspection parameters corresponding to the ultrasonic flaw detection equipment are manually set and cannot be dynamically adjusted according to the relevant parameters of the seamless oil casing, making it difficult to adapt to the actual situation, resulting in poor signal quality and reduced accuracy of the detection results, thereby increasing the risk of misjudgment and missed judgment; in addition, when performing defect analysis, the above-mentioned technology does not describe in detail how to extract key features from ultrasonic signals and the specific meaning of the key features, which will lead to limitations in the actual application of the above-mentioned technology and also lead to inaccurate defect analysis results.

[0005] In view of this, the present invention proposes a non-destructive ultrasonic flaw detection method and system for seamless oil casing to solve the above problems. Summary of the invention

[0006] In order to overcome the above-mentioned defects of the prior art and achieve the above-mentioned purpose, the present invention provides the following technical solution: a non-destructive ultrasonic flaw detection method for seamless oil casing, comprising:

[0007] Collect casing characteristic data;

[0008] Analyze the casing characteristic data and dynamically adjust the ultrasonic parameters;

[0009] Collect ultrasonic flaw detection data;

[0010] Processing ultrasonic flaw detection data to obtain ultrasonic processing data;

[0011] Performing feature extraction on the ultrasonic processing data to extract ultrasonic feature data;

[0012] Analyze ultrasonic feature data to identify casing defects.

[0013] Furthermore, the casing characteristic data includes the wall thickness, casing material and surface roughness;

[0014] The method for collecting the pipe wall thickness is as follows: using an eddy current sensor to collect eddy current signals; presetting calibration coefficients c and d; calculating the pipe wall thickness through the eddy current signals and the calibration coefficients c and d; the expression for the pipe wall thickness is: D=cg+d; where D is the pipe wall thickness and g is the eddy current signal;

[0015] The surface roughness collection method is: use a laser scanner to irradiate the laser beam on the surface of the oil casing for scanning, obtain the height data of a points, a is an integer greater than 1, calculate the height difference data of every two adjacent points in turn, and take the average of all height difference data as the surface roughness; the expression of surface roughness is: In the formula, R is the surface roughness, H is b is the height data of the b-th point, b∈[1,a].

[0016] Furthermore, the step of dynamically adjusting the ultrasound parameters includes:

[0017] Step 1: Define the frequency range [h min ,h max ], loudness extreme value k max , pulse emission rate range [u min ,u max ], loudness attenuation coefficient α and iteration threshold y;

[0018] Step 2: construct m groups of ultrasound parameters, set different digital labels for the m groups of ultrasound parameters, and mark them as ultrasound labels;

[0019] Step 3: Construct a population S, which includes n individuals. The position of each individual is defined in a one-dimensional search space. The range of the one-dimensional search space is m groups of ultrasound parameters. The individual position corresponds to the ultrasound label one by one. The number of iterations t corresponding to the population S is 0.

[0020] Step 4: Determine the matching function;

[0021] Step 5: Generate the migration speed, frequency, loudness and pulse firing rate corresponding to each individual;

[0022] Step 6: Calculate the candidate position of each individual and determine whether to update the position of each individual;

[0023] Step 7: Update the migration speed, frequency, loudness and pulse emission rate corresponding to each individual;

[0024] Step 8: Determine whether the number of iterations t is less than the iteration threshold y. If so, set t=t+1 and return to step 6. If not, proceed to step 9.

[0025] Step 9: Calculate the matching degree of each individual in the population S, obtain the ultrasonic label corresponding to the individual with the maximum matching degree, obtain the corresponding ultrasonic parameters according to the ultrasonic label, and dynamically adjust the ultrasonic flaw detection equipment.

[0026] Furthermore, in step 2, the ultrasonic parameters include gain, delay time and scanning speed; the method for constructing m groups of ultrasonic parameters is: according to the technical parameters of the ultrasonic flaw detection equipment, an ultrasonic range is obtained; the ultrasonic range includes a gain range, a time range and a speed range; the ranges in the ultrasonic range correspond to the parameters in the ultrasonic parameters one by one; a value is randomly selected from each range in the ultrasonic range to construct a group of ultrasonic parameters; a total of m groups of ultrasonic parameters are constructed, and the m groups of ultrasonic parameters are all inconsistent;

[0027] In step 3, the expression of the position of each individual in the population S is: In the formula, is the position of the ith individual, C i is the random coefficient of the ith individual, P i ∈[0,1],i∈[1,n];

[0028] In step 4, the expression of the attraction function is: f=pc; wherein f is the attraction, and pc is the matching degree; the matching degree is calculated by: obtaining the corresponding ultrasound parameters according to the ultrasound labels corresponding to the individual positions, and marking them as real-time parameters; using the real-time parameters and the casing feature data as analysis data, inputting the analysis data into the trained matching analysis model, and predicting the corresponding matching degree;

[0029] The training process of the matching analysis model includes:

[0030] Collect p groups of analysis data in advance, set corresponding matching degrees for all the p groups of analysis data, where p is an integer greater than 1, and convert the analysis data and the corresponding matching degrees into a corresponding set of feature vectors; use each set of feature vectors as the input of a matching analysis model, and the matching analysis model uses a set of predicted matching degrees corresponding to each group of analysis data as output, and uses the actual matching degree corresponding to each group of analysis data as a prediction target, where the actual matching degree is the pre-set matching degree corresponding to the analysis data; minimize the sum of prediction errors of all analysis data as a training target; train the matching analysis model until the sum of prediction errors converges and the training is stopped; the matching analysis model is a deep neural network model.

[0031] Furthermore, in step 5, the expression of the pulse emission rate corresponding to each individual is: ; In the formula, is the pulse emission rate of the ith individual, u min is the minimum value of the pulse emission rate range, u max is the maximum value of the pulse emission rate range, C1 is a random number between [0, 1];

[0032] The expression of the frequency corresponding to each individual is: ; In the formula, is the frequency of the ith individual, h min is the minimum value of the frequency range, h max is the maximum value of the frequency range;

[0033] The expression of the migration speed of each individual is: In the formula, is the migration speed of the ith individual, is the optimal position of the i-th individual, and the position corresponding to the maximum matching degree of the optimal position individual during the migration process;

[0034] The expression of the loudness of each individual is: ; In the formula, is the loudness of the ith individual;

[0035] In step 6, the method for calculating the candidate position of each individual includes:

[0036]

[0037] In the formula, is the candidate position of the i-th individual, e is a natural constant;

[0038] Methods for determining whether to update the location of each individual include:

[0039] Calculate the matching degree corresponding to the position of each individual and mark it as the first matching degree; calculate the matching degree corresponding to the candidate position of each individual and mark it as the second matching degree; compare the first matching degree with the second matching degree; if the first matching degree is greater than or equal to the corresponding second matching degree, do not update the position of the corresponding individual; if the first matching degree is less than the corresponding second matching degree, update the position of the corresponding individual to the corresponding candidate position;

[0040] In step 7, the expression of the pulse emission rate corresponding to each individual after updating is: Where N(0,σ) is a random number in Gaussian noise with mean 0 and standard deviation σ;

[0041] The expression of the migration speed of each individual after updating is:

[0042] The expression of the frequency corresponding to each individual after updating is:

[0043] The expression of the loudness of each individual after updating is:

[0044] Furthermore, the ultrasonic flaw detection data is an echo signal; and the method for obtaining ultrasonic processing data includes:

[0045] Performing denoising processing on ultrasonic flaw detection data to obtain ultrasonic denoised data; performing signal smoothing processing on ultrasonic denoised data to obtain ultrasonic processed data;

[0046] The method for denoising ultrasonic flaw detection data comprises:

[0047] The ultrasonic flaw detection data is digitized to obtain ultrasonic digital data, which includes r″ signal values, where r″ is an integer greater than 1; the preset window size is 2w+1; according to the window size, the denoised signal value corresponding to each signal value in the ultrasonic digital data is calculated; the expression of the denoised signal value is: Wherein, y′(r) is the rth denoised signal value, x′(r+r′) is the r+r′th signal value in the ultrasonic digital data, r′∈[-w,w], r∈[1,r″]; all denoised signal values ​​are combined to form ultrasonic denoised data.

[0048] Furthermore, the method for performing signal smoothing processing on ultrasonic denoising data includes:

[0049]

[0050] Where y″(r) is the rth processed signal value in the ultrasonic processed data, y′(r+r′) is the r+r′th denoised signal value in the ultrasonic denoised data, G(r′) is the weight corresponding to the rth processed signal value in the ultrasonic processed data, and ψ is the smoothing strength.

[0051] Furthermore, the method for extracting ultrasonic feature data includes:

[0052] Each processed signal value in the ultrasonic processed data is transformed to obtain the transformed signal value corresponding to each processed signal value; the expression of the transformed signal value is: y″′(r)=IDFT{-j×DFT{y″(r)}}; wherein y″′(r) is the transformed signal value corresponding to the rth processed signal value, IDFT is discrete Fourier transform, j is an imaginary unit, and DFT is inverse discrete Fourier transform; each processed signal value and the corresponding transformed signal value are analyzed to calculate the instantaneous amplitude, instantaneous phase and instantaneous frequency corresponding to each processed signal value; all instantaneous amplitudes, instantaneous phases and instantaneous frequencies are used as ultrasonic feature data; the expression of the instantaneous amplitude is: Where A(r) is the instantaneous amplitude corresponding to the rth processed signal value; the expression of the instantaneous phase is: Where B(r) is the instantaneous phase corresponding to the rth processed signal value; the expression of instantaneous frequency is: Where C(r) is the instantaneous frequency corresponding to the rth processed signal value, is the time derivative of the instantaneous phase.

[0053] Furthermore, the casing defects include defect type and defect degree;

[0054] The steps to identify casing defects include:

[0055] Step a: Establish fuzzy sets, and divide each type of data in the ultrasonic feature data into multiple fuzzy sets;

[0056] Step b: converting each data in the ultrasonic feature data into the membership degree of each corresponding fuzzy set by fuzzification technology;

[0057] Step c: define fuzzy rules;

[0058] Step d: matching the fuzzified ultrasonic feature data with the fuzzy rules, performing fuzzy reasoning, and obtaining fuzzy reasoning results, the fuzzy reasoning results being the membership degree corresponding to each defect type and the membership degree corresponding to each degree of defect degree;

[0059] Step e: Identify casing defects based on fuzzy reasoning results.

[0060] A nondestructive ultrasonic flaw detection system for seamless oil casing, implementing the nondestructive ultrasonic flaw detection method for seamless oil casing, comprising:

[0061] A data acquisition module, used for collecting casing characteristic data;

[0062] Parameter adjustment module, used to analyze the casing characteristic data and dynamically adjust the ultrasonic parameters;

[0063] A data collection module, used for collecting ultrasonic flaw detection data;

[0064] A data processing module is used to process ultrasonic flaw detection data and obtain ultrasonic processing data;

[0065] A feature extraction module is used to extract features from ultrasonic processing data and extract ultrasonic feature data;

[0066] The defect recognition module is used to analyze ultrasonic feature data and identify casing defects.

[0067] The technical effects and advantages of the nondestructive ultrasonic flaw detection method and system for seamless oil casing of the present invention are as follows:

[0068] According to the casing feature data collected in real time, the ultrasonic parameters are dynamically adjusted to adapt to the dynamic environment, thereby effectively improving the signal quality and stability during ultrasonic flaw detection; ultrasonic flaw detection data is collected based on ultrasonic parameters, and ultrasonic feature data is extracted using signal processing technology, and the defect type and degree in the casing are automatically identified using intelligent algorithms; it can effectively identify defect problems in the casing, reduce the risk of misjudgment and missed judgment, and significantly improve the automation, accuracy and efficiency of non-destructive defect detection, providing reliable protection for the full life cycle management of seamless casing. BRIEF DESCRIPTION OF THE DRAWINGS

[0069] Figure 1 This is a schematic diagram of a nondestructive ultrasonic flaw detection system for seamless oil casing according to Example 1 of the present invention;

[0070] Figure 2 This is a flow chart of a method for dynamically adjusting ultrasonic parameters according to Embodiment 1 of the present invention;

[0071] Figure 3 This is a flow chart of a nondestructive ultrasonic flaw detection method for seamless oil casing according to Example 2 of the present invention. DETAILED DESCRIPTION

[0072] The following will be combined with the drawings in the embodiments of the present invention to clearly and completely describe the technical solutions in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments of the present invention, all other embodiments obtained by ordinary technicians in this field without creative work are within the scope of protection of the present invention.

[0073] Example 1

[0074] See also Figure 1 As shown, the nondestructive ultrasonic flaw detection system for seamless oil casing described in this embodiment includes a data acquisition module, a parameter adjustment module, a data collection module, a data processing module, a feature extraction module and a defect recognition module; each module is connected by wire and / or wireless means to realize data transmission between modules;

[0075] The data acquisition module is used to collect casing characteristic data.

[0076] Casing characteristic data include wall thickness, casing material and surface roughness;

[0077] The method for collecting pipe wall thickness is: using an eddy current sensor to collect eddy current signals, the eddy current sensor is integrated on the ultrasonic flaw detection equipment, and the ultrasonic flaw detection equipment is an ultrasonic flaw detector; preset calibration coefficients c and d, the calibration coefficients are determined by technical personnel in the field through experimental fitting during the historical pipe wall thickness measurement process; the pipe wall thickness is calculated through the eddy current signal and the calibration coefficients c and d; the expression of the pipe wall thickness is: D=cg+d; where D is the pipe wall thickness and g is the eddy current signal.

[0078] The casing material is obtained by technicians in this field by testing the oil casing.

[0079] The surface roughness collection method is: using a laser scanner, the laser scanner is integrated on the ultrasonic flaw detection equipment; irradiating the laser beam on the surface of the oil casing for scanning, obtaining the height data of a points, a is an integer greater than 1, calculating the height difference data of every two adjacent points in turn, and taking the average of all height difference data as the surface roughness; the expression of surface roughness is: In the formula, R is the surface roughness, H is b is the height data of the b-th point, b∈[1,a].

[0080] The thickness of the pipe wall directly affects the propagation time and attenuation of the ultrasonic wave. The thicker the pipe wall is, the greater the gain and delay time required to ensure that the ultrasonic signal can be accurately received, and vice versa. Different oil casing materials have different sound speeds and attenuation characteristics, so the gain and delay time that need to be set are different. The surface roughness of the oil casing will affect the incidence and reflection effects of the ultrasonic wave. The rougher the surface of the oil casing, the greater the signal attenuation and distortion. Therefore, the required gain should be greater to ensure that the echo signal of sufficient strength is detected, and the required scanning speed should be smaller, so that the ultrasonic wave has more time to propagate and reflect on the surface of the oil casing, which is conducive to the collection and analysis of the echo signal to obtain more reliable defect detection results.

[0081] The parameter adjustment module is used to analyze the casing characteristic data and dynamically adjust the ultrasonic parameters.

[0082] like Figure 2 As shown, the steps of dynamically adjusting ultrasound parameters include:

[0083] Step 1: Define the frequency range [h min ,h max ], loudness extreme value k max , pulse emission rate range [u min ,u max ], loudness attenuation coefficient α and iteration threshold y;

[0084] Step 2: construct m groups of ultrasound parameters, set different digital labels for the m groups of ultrasound parameters, and mark them as ultrasound labels;

[0085] Step 3: Construct a population S, which includes n individuals. The position of each individual is defined in a one-dimensional search space. The range of the one-dimensional search space is m groups of ultrasound parameters. The individual position corresponds to the ultrasound label one by one. The number of iterations t corresponding to the population S is 0.

[0086] Step 4: Determine the matching function;

[0087] Step 5: Generate the migration speed, frequency, loudness and pulse firing rate corresponding to each individual;

[0088] Step 6: Calculate the candidate position of each individual and determine whether to update the position of each individual;

[0089] Step 7: Update the migration speed, frequency, loudness and pulse emission rate corresponding to each individual;

[0090] Step 8: Determine whether the number of iterations t is less than the iteration threshold y. If so, set t=t+1 and return to step 6. If not, proceed to step 9.

[0091] Step 9: Calculate the matching degree of each individual in the population S, obtain the ultrasonic label corresponding to the individual with the maximum matching degree, obtain the corresponding ultrasonic parameters according to the ultrasonic label, and dynamically adjust the ultrasonic flaw detection equipment.

[0092] In the above step 1, the frequency range [h min ,h max ], loudness extreme value k max , pulse emission rate range [u min ,u max ], the loudness attenuation coefficient α and the iteration threshold y are both pre-defined by those skilled in the art according to actual conditions.

[0093] In the above step 2, the ultrasonic parameters include gain, delay time and scanning speed; wherein gain is the ability to amplify the strength or amplitude of the ultrasonic signal; delay time is the time interval between transmitting the ultrasonic wave and receiving the echo signal; scanning speed is the moving speed of the ultrasonic flaw detection equipment on the surface of the oil casing;

[0094] The method for constructing m groups of ultrasonic parameters is as follows: according to the technical parameters of the ultrasonic flaw detection equipment, an ultrasonic range is obtained; the ultrasonic range includes a gain range, a time range and a speed range; the ranges in the ultrasonic range correspond to the parameters in the ultrasonic parameters one by one; a value is randomly selected from each range in the ultrasonic range to construct a group of ultrasonic parameters; a total of m groups of ultrasonic parameters are constructed, and the m groups of ultrasonic parameters are all inconsistent.

[0095] In step 3 above, the expression for the position of each individual in population S is: In the formula, is the position of the ith individual, C i is the random coefficient of the ith individual, P i ∈[0,1],i∈[1,n].

[0096] In the above step 4, the expression of the attraction function is: f=pc; where f is the attraction and pc is the matching degree; the matching degree is calculated as follows: according to the ultrasound label corresponding to the individual position, the corresponding ultrasound parameters are obtained and marked as real-time parameters; the real-time parameters and the casing feature data are used as analysis data, and the analysis data is input into the trained matching analysis model to predict the corresponding matching degree.

[0097] The specific training process of the matching analysis model includes:

[0098] Collect p groups of analysis data in advance, set corresponding matching degrees for the p groups of analysis data, where p is an integer greater than 1, and convert the analysis data and the corresponding matching degrees into a corresponding set of feature vectors; the matching degrees corresponding to the analysis data are collected by a technician in the field during the dynamic adjustment of historical ultrasonic parameters, and each group of analysis data is analyzed in turn in combination with actual experience, and the degree of adaptation between the real-time parameters and the cannula characteristic data in each group of analysis data is evaluated, and the degree of adaptation between the real-time parameters and the cannula characteristic data in each group of analysis data is used as the matching degree of the corresponding analysis data, and the corresponding matching degrees are set for the p groups of analysis data in turn;

[0099] Each set of feature vectors is used as the input of the matching analysis model. The matching analysis model takes a set of predicted matching degrees corresponding to each set of analysis data as output, and takes the actual matching degree corresponding to each set of analysis data as the prediction target. The actual matching degree is the pre-set matching degree corresponding to the analysis data. The training goal is to minimize the sum of the prediction errors of all analysis data. The calculation formula of the prediction error is η q =(ε q -γ q ) 2 , where η q is the prediction error, q is the group number of the eigenvector corresponding to the analyzed data, ε q is the predicted matching degree corresponding to the qth group of analysis data, γ q is the actual matching degree corresponding to the qth group of analysis data; the matching analysis model is trained until the sum of the prediction errors reaches convergence and the training is stopped.

[0100] The above matching analysis model is specifically a deep neural network model; it includes an input layer, a hidden layer and an output layer; each hidden layer includes multiple neurons, each neuron is connected to the neurons in the next layer, and the connection contains weights, which determine the importance and influence of data transmission in the neural network; an activation function is applied to each neuron between the hidden layer and the output layer, and the activation function introduces nonlinearity, allowing the network to learn more complex patterns and features.

[0101] In step 5 above, the expression of the pulse emission rate corresponding to each individual is: In the formula, is the pulse emission rate of the ith individual, u min is the minimum value of the pulse emission rate range, u max is the maximum value of the pulse emission rate range, C1 is a random number between [0, 1];

[0102] The expression of the frequency corresponding to each individual is: ; In the formula, is the frequency of the ith individual, h min is the minimum value of the frequency range, h max is the maximum value of the frequency range;

[0103] The expression of the migration speed of each individual is: In the formula, is the migration speed of the ith individual, is the optimal position of the i-th individual, and the optimal position is the position corresponding to the maximum matching degree of the individual during the migration process;

[0104] The expression of the loudness of each individual is: ; In the formula, is the loudness of the ith individual.

[0105] In the above step 6, the method for calculating the candidate position of each individual includes:

[0106]

[0107] In the formula, is the candidate position of the i-th individual, and e is a natural constant.

[0108] Methods for determining whether to update the location of each individual include:

[0109] Calculate the matching degree corresponding to the position of each individual and mark it as the first matching degree; calculate the matching degree corresponding to the candidate position of each individual and mark it as the second matching degree; compare the first matching degree with the second matching degree; if the first matching degree is greater than or equal to the corresponding second matching degree, do not update the position of the corresponding individual; if the first matching degree is less than the corresponding second matching degree, update the corresponding individual position to the corresponding candidate position.

[0110] In the above step 7, the expression of the pulse emission rate corresponding to each individual after update is: Wherein, N(0,σ) is a random number in Gaussian noise with a mean of 0 and a standard deviation of σ. The standard deviation σ of the Gaussian noise is set by those skilled in the art according to the required degree of disturbance;

[0111] The expression of the migration speed of each individual after updating is:

[0112] The expression of the frequency corresponding to each individual after updating is:

[0113] The expression of the loudness of each individual after updating is: .

[0114] The data collection module is used to collect ultrasonic flaw detection data.

[0115] The ultrasonic flaw detection data is an echo signal, which is a signal generated by the reflection of ultrasonic waves sent by the ultrasonic flaw detection equipment to the surface of the oil casing. The ultrasonic flaw detection data is obtained by the ultrasonic flaw detection equipment. The echo signal contains key information about the defects of the oil casing. By analyzing the echo signal, the quality and safety of the oil casing can be identified and evaluated. It should be understood that the ultrasonic flaw detection equipment is a movable device that can continuously scan and detect along the surface of the oil casing to perform all-round detection of the oil casing to ensure the integrity of the defect detection.

[0116] The data processing module is used to process the ultrasonic flaw detection data and obtain ultrasonic processing data.

[0117] Methods for acquiring sonication data include:

[0118] The ultrasonic flaw detection data is subjected to denoising to obtain ultrasonic denoising data; and the ultrasonic denoising data is subjected to signal smoothing to obtain ultrasonic processing data.

[0119] The methods for denoising ultrasonic flaw detection data include:

[0120] The ultrasonic flaw detection data is digitized to obtain ultrasonic digital data, wherein the ultrasonic digital data includes r″ signal values, where r″ is an integer greater than 1; the digitization process is a prior art and will not be described in detail herein; the preset window size is 2w+1, and the window size is preset by those skilled in the art according to actual conditions; according to the window size, the denoised signal value corresponding to each signal value in the ultrasonic digital data is calculated; the expression of the denoised signal value is: Wherein, y′(r) is the rth denoised signal value, x′(r+r′) is the r+r′th signal value in the ultrasonic digital data, r′∈[-w,w], r∈[1,r″]; all denoised signal values ​​are combined to form ultrasonic denoised data.

[0121] The method for signal smoothing of ultrasonic denoising data includes:

[0122]

[0123] Wherein, y″(r) is the rth processed signal value in the ultrasonic processed data, y′(r+r′) is the r+r′th denoised signal value in the ultrasonic denoised data, G(r′) is the weight corresponding to the rth processed signal value in the ultrasonic processed data, indicating the contribution of the r+r′th denoised signal value in the ultrasonic denoised data to the rth processed signal value in the ultrasonic processed data, ψ is the smoothing strength, and the smoothing strength is set by those skilled in the art according to actual conditions.

[0124] The feature extraction module is used to extract features from ultrasonic processing data and extract ultrasonic feature data.

[0125] Methods for extracting ultrasonic feature data include:

[0126] Each processed signal value in the ultrasonic processed data is transformed to obtain the transformed signal value corresponding to each processed signal value; the expression of the transformed signal value is: y″′(r)=IDFT{-j×DFT{y″(r)}}; wherein y″′(r) is the transformed signal value corresponding to the rth processed signal value, IDFT is discrete Fourier transform, j is an imaginary unit, and DFT is inverse discrete Fourier transform; wherein discrete Fourier transform and inverse discrete Fourier transform are both prior arts and will not be described in detail here; each processed signal value and the corresponding transformed signal value are analyzed to calculate the instantaneous amplitude, instantaneous phase and instantaneous frequency corresponding to each processed signal value; all instantaneous amplitudes, instantaneous phases and instantaneous frequencies are used as ultrasonic feature data; the expression of instantaneous amplitude is: Where A(r) is the instantaneous amplitude corresponding to the rth processed signal value; the expression of the instantaneous phase is: Where B(r) is the instantaneous phase corresponding to the rth processed signal value; the expression of instantaneous frequency is: Where C(r) is the instantaneous frequency corresponding to the rth processed signal value, is the time derivative of the instantaneous phase.

[0127] The defect recognition module is used to analyze ultrasonic feature data and identify casing defects.

[0128] Casing defects include defect type and defect degree; defect type includes cracks, corrosion, etc.; defect degree includes severe, moderate and slight.

[0129] The steps to identify casing defects include:

[0130] Step a: Establish fuzzy sets, and divide each type of data in the ultrasonic feature data into multiple fuzzy sets; for example, the fuzzy sets corresponding to the instantaneous amplitude are low amplitude, medium amplitude, high amplitude, etc., the fuzzy sets corresponding to the instantaneous phase are large phase change, medium phase change, small phase change, etc., and the fuzzy sets corresponding to the instantaneous frequency are large frequency change, medium frequency change, small frequency change, etc.;

[0131] Step b: convert each data in the ultrasonic feature data into the membership of each corresponding fuzzy set through fuzzification technology; fuzzification technology is the process of converting an exact value into the membership corresponding to a fuzzy set, and fuzzification technology includes, for example, a triangular membership function, a trapezoidal membership function, etc.; for example, the value of the instantaneous amplitude is low, and it is inferred that the membership of the low amplitude is 0.9, the membership of the medium amplitude is 0.1, and the membership of the high amplitude is 0;

[0132] Step c: define fuzzy rules, which are defined based on expert knowledge or relevant literature; for example, if the amplitude is high, the phase change is large, and the frequency change is large, then the defect type is inferred to be crack and the defect degree is severe; if the amplitude is medium, the phase change is large, and the frequency change is medium, then the defect type is inferred to be corrosion and the defect degree is medium; if the amplitude is low, the phase change is small, and the frequency change is small, then the defect type is inferred to be non-defective;

[0133] Step d: Match the fuzzified ultrasonic feature data with the fuzzy rules, perform fuzzy reasoning, and obtain the fuzzy reasoning results, which are the membership corresponding to each type of defect type and the membership corresponding to each degree of defect degree; the fuzzy reasoning method is, for example, the Mamdani or Sugeno fuzzy reasoning method; the fuzzy reasoning results are, for example, the membership of corrosion is 0.3, the membership of crack is 0.7, and the membership of normal is 0; the membership of severe is 0.8, the membership of moderate is 0.2, and the membership of slight is 0;

[0134] Step e: Identify casing defects based on fuzzy reasoning results.

[0135] This embodiment dynamically adjusts ultrasonic parameters according to the casing characteristic data collected in real time to adapt to the dynamic environment, thereby effectively improving the signal quality and stability during ultrasonic flaw detection; collects ultrasonic flaw detection data based on ultrasonic parameters, extracts ultrasonic characteristic data using signal processing technology, and automatically identifies the defect type and degree in the oil casing using an intelligent algorithm; can effectively identify defect problems in the oil casing, reduce the risk of misjudgment and missed judgment, significantly improve the automation level, accuracy and efficiency of non-destructive defect detection, and provide reliable protection for the full life cycle management of seamless oil casing.

[0136] Example 2

[0137] See also Figure 3 As shown, the part not described in detail in this embodiment is described in Example 1, and a non-destructive ultrasonic flaw detection method for seamless oil casing is provided, the method comprising:

[0138] Collect casing characteristic data;

[0139] Analyze the casing characteristic data and dynamically adjust the ultrasonic parameters;

[0140] Collect ultrasonic flaw detection data;

[0141] Processing ultrasonic flaw detection data to obtain ultrasonic processing data;

[0142] Performing feature extraction on the ultrasonic processing data to extract ultrasonic feature data;

[0143] Analyze ultrasonic feature data to identify casing defects.

[0144] Example 3

[0145] The present application also provides an electronic device. The electronic device may include one or more processors and one or more memories. The memories store computer-readable codes, and when the computer-readable codes are run by the one or more processors, the non-destructive ultrasonic flaw detection method for seamless oil casing as described above may be executed.

[0146] The method or system according to the implementation mode of the present application can also be implemented with the help of the architecture of the electronic device shown in the present application. The electronic device may include a bus, one or more CPUs, ROM, RAM, a communication port connected to a network, input / output, a hard disk, etc. A storage device in the electronic device, such as a ROM or a hard disk, can store a non-destructive ultrasonic flaw detection method for seamless oil casing provided in the present application. Furthermore, the electronic device may also include a user interface. Of course, the architecture shown in the present application is only exemplary. When implementing different devices, one or more components in the electronic device shown in the present application may be omitted according to actual needs.

[0147] Example 4

[0148] One embodiment of the present application discloses a computer-readable storage medium. Computer-readable instructions are stored on the computer-readable storage medium. When the computer-readable instructions are executed by a processor, a non-destructive ultrasonic flaw detection method for seamless oil casing according to an embodiment of the present application described with reference to the above figures can be executed. The storage medium includes, but is not limited to, for example, volatile memory and / or non-volatile memory. Volatile memory may include, for example, random access memory (RAM) and cache memory (cache). Non-volatile memory may include, for example, read-only memory (ROM), hard disk, flash memory, etc.

[0149] In addition, according to the embodiments of the present application, the process described above with reference to the flowchart can be implemented as a computer software program. For example, the present application provides a non-transitory machine-readable storage medium, which stores machine-readable instructions, and the machine-readable instructions can be executed by a processor to execute instructions corresponding to the method steps provided in the present application, for example: a non-destructive ultrasonic flaw detection method for seamless oil casing. When the computer program is executed by a central processing unit (CPU), the above functions defined in the method of the present application are executed.

[0150] The above is only a specific embodiment of the present invention, but the protection scope of the present invention is not limited thereto. Any person skilled in the art who is familiar with the technical field can easily think of changes or substitutions within the technical scope disclosed by the present invention, which should be included in the protection scope of the present invention. Therefore, the protection scope of the present invention should be based on the protection scope of the claims.

[0151] Finally: The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. Any modifications, equivalent substitutions, improvements, etc. made within the spirit and principles of the present invention should be included in the protection scope of the present invention.

Claims

1. A nondestructive ultrasonic flaw detection method for seamless oil casing, characterized in that: include: Collect casing characteristic data; Analyze the casing characteristic data and dynamically adjust the ultrasonic parameters; Collect ultrasonic flaw detection data; Processing ultrasonic flaw detection data to obtain ultrasonic processing data; Performing feature extraction on the ultrasonic processing data to extract ultrasonic feature data; Analyze ultrasonic feature data to identify casing defects.

2. A nondestructive ultrasonic flaw detection method for seamless oil casing according to claim 1, characterized in that: The casing characteristic data includes the wall thickness, casing material and surface roughness; The method for collecting the pipe wall thickness is as follows: using an eddy current sensor to collect eddy current signals; presetting calibration coefficients c and d; calculating the pipe wall thickness through the eddy current signals and the calibration coefficients c and d; the expression for the pipe wall thickness is: D=cg+d; where D is the pipe wall thickness and g is the eddy current signal; The surface roughness collection method is: use a laser scanner to irradiate the laser beam on the surface of the oil casing for scanning, obtain the height data of a points, a is an integer greater than 1, calculate the height difference data of every two adjacent points in turn, and take the average of all height difference data as the surface roughness; the expression of surface roughness is: In the formula, R is the surface roughness, H is b is the height data of the b-th point, b∈[1,a].

3. A nondestructive ultrasonic flaw detection method for seamless oil casing according to claim 2, characterized in that: The step of dynamically adjusting the ultrasound parameters comprises: Step 1: Define the frequency range [h min ,h max ], loudness extreme value k max , pulse emission rate range [u min ,u max ], loudness attenuation coefficient α and iteration threshold y; Step 2: construct m groups of ultrasound parameters, set different digital labels for the m groups of ultrasound parameters, and mark them as ultrasound labels; Step 3: Construct a population S, which includes n individuals. The position of each individual is defined in a one-dimensional search space. The range of the one-dimensional search space is m groups of ultrasound parameters. The individual position corresponds to the ultrasound label one by one. The number of iterations t corresponding to the population S is 0. Step 4: Determine the matching function; Step 5: Generate the migration speed, frequency, loudness and pulse firing rate corresponding to each individual; Step 6: Calculate the candidate position of each individual and determine whether to update the position of each individual; Step 7: Update the migration speed, frequency, loudness and pulse emission rate corresponding to each individual; Step 8: Determine whether the number of iterations t is less than the iteration threshold y. If so, set t=t+1 and return to step 6. If not, proceed to step 9. Step 9: Calculate the matching degree of each individual in the population S, obtain the ultrasonic label corresponding to the individual with the maximum matching degree, obtain the corresponding ultrasonic parameters according to the ultrasonic label, and dynamically adjust the ultrasonic flaw detection equipment.

4. A nondestructive ultrasonic flaw detection method for seamless oil casing according to claim 3, characterized in that: In step 2, the ultrasonic parameters include gain, delay time and scanning speed; the method for constructing m groups of ultrasonic parameters is: according to the technical parameters of the ultrasonic flaw detection equipment, an ultrasonic range is obtained; the ultrasonic range includes a gain range, a time range and a speed range; the ranges in the ultrasonic range correspond to the parameters in the ultrasonic parameters one by one; a value is randomly selected from each range in the ultrasonic range to construct a group of ultrasonic parameters; a total of m groups of ultrasonic parameters are constructed, and the m groups of ultrasonic parameters are all inconsistent; In step 3, the expression of the position of each individual in the population S is: In the formula, is the position of the ith individual, C i is the random coefficient of the ith individual, P i ∈[0,1],i∈[1,n]; In step 4, the expression of the attraction function is: f=pc; Where f is the attraction, pc is the matching degree; the matching degree is calculated as follows: according to the ultrasound label corresponding to the individual position, the corresponding ultrasound parameters are obtained and marked as real-time parameters; the real-time parameters and the casing feature data are used as analysis data, and the analysis data is input into the trained matching analysis model to predict the corresponding matching degree; The training process of the matching analysis model includes: Collect p groups of analysis data in advance, set corresponding matching degrees for all the p groups of analysis data, where p is an integer greater than 1, and convert the analysis data and the corresponding matching degrees into a corresponding set of feature vectors; use each set of feature vectors as input to a matching analysis model, which outputs a set of predicted matching degrees corresponding to each group of analysis data, and uses the actual matching degree corresponding to each group of analysis data as a prediction target, where the actual matching degree is the pre-set matching degree corresponding to the analysis data; and use minimizing the sum of the prediction errors of all analysis data as a training target; The matching analysis model is trained until the sum of the prediction errors reaches convergence and then the training is stopped; the matching analysis model is a deep neural network model.

5. A nondestructive ultrasonic flaw detection method for seamless oil casing according to claim 4, characterized in that: In step 5, the expression of the pulse emission rate corresponding to each individual is: In the formula, is the pulse emission rate of the ith individual, u min is the minimum value of the pulse emission rate range, u max is the maximum value of the pulse emission rate range, C1 is a random number between [0, 1]; The expression of the frequency corresponding to each individual is: In the formula, is the frequency of the ith individual, h min is the minimum value of the frequency range, h max is the maximum value of the frequency range; The expression of the migration speed of each individual is: In the formula, is the migration speed of the ith individual, is the optimal position of the i-th individual, and the position corresponding to the maximum matching degree of the optimal position individual during the migration process; The expression of the loudness of each individual is: In the formula, is the loudness of the ith individual; In step 6, the method for calculating the candidate position of each individual includes: In the formula, is the candidate position of the ith individual, e is a natural constant; Methods for determining whether to update the location of each individual include: Calculate the matching degree corresponding to the position of each individual and mark it as the first matching degree; calculate the matching degree corresponding to the candidate position of each individual and mark it as the second matching degree; compare the first matching degree with the second matching degree; if the first matching degree is greater than or equal to the corresponding second matching degree, do not update the position of the corresponding individual; if the first matching degree is less than the corresponding second matching degree, update the position of the corresponding individual to the corresponding candidate position; In step 7, the expression of the pulse emission rate corresponding to each individual after updating is: Where N(0,σ) is a random number in Gaussian noise with mean 0 and standard deviation σ; The expression of the migration speed of each individual after updating is: The expression of the frequency corresponding to each individual after updating is: The expression of the loudness of each individual after updating is:

6. A nondestructive ultrasonic flaw detection method for seamless oil casing according to claim 5, characterized in that: The ultrasonic flaw detection data is an echo signal; the method for obtaining ultrasonic processing data includes: Performing denoising processing on ultrasonic flaw detection data to obtain ultrasonic denoised data; performing signal smoothing processing on ultrasonic denoised data to obtain ultrasonic processed data; The method for denoising ultrasonic flaw detection data comprises: The ultrasonic flaw detection data is digitized to obtain ultrasonic digital data, which includes r″ signal values, where r″ is an integer greater than 1; the preset window size is 2w+1; according to the window size, the denoised signal value corresponding to each signal value in the ultrasonic digital data is calculated; the expression of the denoised signal value is: Wherein, y′(r) is the rth denoised signal value, x′(r+r′) is the r+r′th signal value in the ultrasonic digital data, r′∈[-w,w], r∈[1,r″]; all denoised signal values ​​are combined to form ultrasonic denoised data.

7. A nondestructive ultrasonic flaw detection method for seamless oil casing according to claim 6, characterized in that: The method for performing signal smoothing processing on ultrasonic denoising data comprises: Where y″(r) is the rth processed signal value in the ultrasonic processed data, y′(r+r′) is the r+r′th denoised signal value in the ultrasonic denoised data, G(r′) is the weight corresponding to the rth processed signal value in the ultrasonic processed data, and ψ is the smoothing strength.

8. A nondestructive ultrasonic flaw detection method for seamless oil casing according to claim 7, characterized in that: The method for extracting ultrasonic feature data comprises: Each processed signal value in the ultrasonic processed data is transformed to obtain the transformed signal value corresponding to each processed signal value; the expression of the transformed signal value is: y″′(r)=IDFT{-j×DFT{y″(r)}}; wherein y″′(r) is the transformed signal value corresponding to the rth processed signal value, IDFT is discrete Fourier transform, j is an imaginary unit, and DFT is inverse discrete Fourier transform; each processed signal value and the corresponding transformed signal value are analyzed to calculate the instantaneous amplitude, instantaneous phase and instantaneous frequency corresponding to each processed signal value; all instantaneous amplitudes, instantaneous phases and instantaneous frequencies are used as ultrasonic feature data; the expression of the instantaneous amplitude is: Where A(r) is the instantaneous amplitude corresponding to the rth processed signal value; the expression of the instantaneous phase is: Where B(r) is the instantaneous phase corresponding to the rth processed signal value; the expression of instantaneous frequency is: Where C(r) is the instantaneous frequency corresponding to the rth processed signal value, is the time derivative of the instantaneous phase.

9. A nondestructive ultrasonic flaw detection method for seamless oil casing according to claim 8, characterized in that: The casing defects include defect type and defect degree; The steps to identify casing defects include: Step a: Establish fuzzy sets, and divide each type of data in the ultrasonic feature data into multiple fuzzy sets; Step b: converting each data in the ultrasonic feature data into the membership degree of each corresponding fuzzy set by fuzzification technology; Step c: define fuzzy rules; Step d: matching the fuzzified ultrasonic feature data with the fuzzy rules, performing fuzzy reasoning, and obtaining fuzzy reasoning results, the fuzzy reasoning results being the membership degree corresponding to each defect type and the membership degree corresponding to each degree of defect degree; Step e: Identify casing defects based on fuzzy reasoning results.

10. A nondestructive ultrasonic flaw detection system for seamless oil casing, implementing a nondestructive ultrasonic flaw detection method for seamless oil casing as claimed in any one of claims 1 to 9, characterized in that: include: A data acquisition module, used for collecting casing characteristic data; Parameter adjustment module, used to analyze the casing characteristic data and dynamically adjust the ultrasonic parameters; A data collection module, used for collecting ultrasonic flaw detection data; A data processing module is used to process ultrasonic flaw detection data and obtain ultrasonic processing data; A feature extraction module is used to extract features from ultrasonic processing data and extract ultrasonic feature data; The defect recognition module is used to analyze ultrasonic feature data and identify casing defects.

Images