Vortex detection defect signal generation method for heat transfer tube of steam generator

Abstract

The invention belongs to the technical field of nondestructive testing, and particularly relates to a steam generator heat transfer tube eddy current testing defect signal generation method. The method comprises the following steps: selecting signal eddy current detection data containing defects, extracting original data points of defect signals and relevant parameters of defect calibration, selecting other eddy current detection data as target data, fusing the defect signals to a certain specified position of the target data to form new defect signals, including defect signal information extraction; aligning the defect signal data point set based on the selected other eddy current detection data; calculating offsets of all data points and a central point of the defect signal; and fusing other target eddy current data. The method has the beneficial effects that a real defect signal is selected, and defect data points and detection parameters are extracted; selecting other eddy current detection data without defect signals; and the defect data is amplified or reduced according to the acquisition speed, so that the length of the fused defect can be kept unchanged.

Description

TECHNICAL FIELD

[0001] The application belongs to the technical field of nondestructive testing, and particularly relates to a steam generator heat pipe eddy current detection defect signal generation method. BACKGROUND

[0002] The steam generator is a key component of a nuclear power facility, and its heat pipe accounts for more than 70% of the one-loop boundary area. Since the wall is thin, defects are prone to occur, which is the focus of nuclear operation safety. The most rapid and effective method for effective operation and maintenance detection is eddy current detection, and the core technology of eddy current detection is signal analysis and defect identification. At present, defect identification mainly relies on analysis experts for identification. It takes 5-8 years to train a qualified signal analyst, and the core is to accumulate a large number of case experience. Since the nuclear power operation time in China is limited, the defect signal sample is seriously lacking, which leads to limited defect types seen by the analyst, insufficient judgment experience, and easy to produce missed judgment or misjudgment. At the same time, with the application of artificial intelligence in various industries, the use of artificial intelligence for defect signal identification has become a development trend, and artificial intelligence is more dependent on defect sample signals.

[0003] At present, the acquisition of defect signals is mainly through the collection of real defects of the evaporator heat pipe and the acquisition of defect signals by making artificial defect test pieces. Artificially made defects are different from natural defects and are expensive. SUMMARY

[0004] The purpose of the present application is to provide a steam generator heat pipe eddy current detection defect signal generation method, which generates defect signals by providing a method based on real defects and real data fusion, and solves the problem of few steam generator heat defect samples.

[0005] The technical scheme of the present application is as follows: a steam generator heat pipe eddy current detection defect signal generation method, selecting defect-containing signal eddy current detection data, extracting defect signal original data points and defect calibration related parameters, selecting other eddy current detection data as target data, and fusing the defect signal to a certain specified position of the target data to form a new defect signal. The specific steps are as follows:

[0006] Step 1: defect signal information extraction;

[0007] Step 2: aligning the defect signal data point set based on the selected other eddy current detection data;

[0008] Step 3: calculating the offset of all data points of the defect signal from the center point;

[0009] Step 4: fusing the other target eddy current data.

[0010] The step 1 comprises:

[0011] Step 11: open the data set containing the defect signal, calibrate the signal according to the calibration tube data, and obtain the relevant calibration parameters: including the rotation angle value of each channel, marked as Rotation; the voltage normalization coefficient of each channel, Voltscale, and by setting the defect-free part in the heat pipe as the center point of the channel, NullPoint, the eddy current signal data points, including two components;

[0012] Step 12: read the eddy current detection data containing the defect signal, display the signal using a long bar chart and a Li Sha Yu chart, select the defect signal, determine the starting point of the defect signal, and select all data points between point A as the starting point and point B as the ending point as the defect signal data points PointNum1, extract the data point set within the starting point range of the current channel to form the defect data point set of the current channel;

[0013] Step 13: calculate the defect tube probe movement speed Speed1, the calculation method: the total length of the defect tube divided by the ratio of the effective data point number in the defect tube eddy current signal tube to the sampling rate;

[0014] Step 14: calculate the defect length, the calculation method: the number of defect starting point data points divided by the sampling rate, then multiplied by the speed Speed1, to obtain the defect length Len1.

[0015] The step 2 comprises:

[0016] Step 21: obtain and calculate the relevant parameters of the data set of other eddy current detection data, including the sampling rate Sample2 and the probe movement speed Speed2;

[0017] Step 22: calculate the data point number of other eddy current detection data corresponding to the defect length Len1, the calculation method: the defect length Len1 divided by the speed Speed2 and then multiplied by the Sample2 to obtain the data point number PointNum2;

[0018] Step 23: when PointNum2 is less than PointNum1, compress the defect data point set to the length equal to PointNum1, when PointNum2 is greater than PointNum1, expand the defect data point set to the length equal to PointNum1, and when they are equal, no processing is performed;

[0019] Step 24: process different compression situations.

[0020] The step 24 comprises:

[0021] Calculate the compression coefficient PointNum1 / PointNum2;

[0022] Create an array DataNew[PointNum2] to store data points for this data segment to be reduced. Each data unit is an eddy current signal data point including two components, X and Y. For the data segment PointNum1 data points Data[PointNum1], perform reduction processing sequentially from the first point to the last point. Assume that the current data point index is I, and its corresponding newly generated data index is J, J = (INT)(I / (PointNum1 / PointNum2)), and the data point is DataNew[J].

[0023] Using the Vmax measurement method, the amplitudes of Data[I] and DataNew[J] to the center point are calculated respectively. Assuming that the amplitude calculated for Data[I] is Vi and the amplitude calculated for DataNew[J] is Vj, Vi and Vj are compared. If Vi is greater than Vj, the following assignment is performed: DataNew[J] = Data[I]; until all m data points have been converted, the DataNew array is used as the compressed data.

[0024] Step 24 includes the following processing for the expanded case:

[0025] Calculate the scaling factor PointNum2 / PointNum1;

[0026] For the data points Data[PointNum1] of the data segment PointNum1, from the first point to the last point, the data points that need to be interpolated between adjacent data points are calculated in turn. The linear interpolation method is used to calculate and interpolate the X and Y components of each data point to form the dataset DataNew[PointNum2] as the data point set of the defect signal.

[0027] Step 3 includes:

[0028] Step 31: Calculate the refNull of the data points after center point phase calibration. Calculation method:

[0029] Using the rotation formula in a Cartesian coordinate system, let the rotation angle of the coordinate axis be θ, and let P be any point in the plane with coordinates (x, y) in the original coordinate system xOy and (x′, y′) in the new coordinate system x′Oy′.

[0030]

[0031] Using the above formula, rotate the center point of the reference channel by the rotation angle Rotation to obtain a new data point, which is marked as refNull;

[0032] Step 32: Calculate the offset of the defect signal data point set relative to the reference center point, forming a deviation value set with the same number of defect signal data points.

[0033] The calculation method of step 32 is as follows:

[0034] Rotate each data point in the defect signal data point set by the rotation angle Rotation, and calculate the rotated data point set.

[0035] Calculate the deviation of each data point in the data point set, the calculation method is: the x of each point minus the x of refNull point, the y of each point minus the y of refNull point, to get the deviation point, save all deviation points to the deviation value data set, marked as PointList.

[0036] The step 4 includes:

[0037] Step 41: Open the eddy current detection data set containing the target data, calibrate the signal according to the calibration tube data, and obtain the relevant calibration parameters: including the rotation angle value of each channel, marked as Rotation2; The voltage normalization coefficient Voltscale2 of each channel and the center point NullPoint2 of the channel by setting the defect-free position in the heat pipe as the center point;

[0038] Step 42: Select the fusion data region of the target data, including the data start point and the data end point, and the data length is PointNum2;

[0039] Step 43: Calculate the voltage normalization coefficient ratio of the target data channel and the defect data channel, marked as belt1:

[0040] Step 44: Fuse the data points in the current target data channel fusion data region.

[0041] The fusion processing of step 44 includes that the target data channel fusion data region range is all data points within the index start2 to end2 range, and the current data point is marked as CurDP, the index is index, and the corresponding defect deviation data set PointList index is index-start2, and the two points are fused.

[0042] The fusion processing of step 44 includes the following:

[0043] After calculating the current data point rotation Rotation2, the data point is obtained, marked as Base;

[0044] PointList index of data point of index-start2, marked as Crack; calculate x1 equal to X of Crack divided by belt1, y1 equal to y of Crack divided by belt1;

[0045] Add x1 to x of CurDP, add y1 to y of CurDP, to obtain data point NextDP;

[0046] The data point obtained by rotating NextDP by Rotation2*(-1) degrees is taken as a newly generated data point after fusion of CurDP;

[0047] All the fused data points are saved into the retList object as new data after fusion of the target channel defect data, and the fusion information contained in the data is the newly generated defect signal.

[0048] The beneficial effects of the present application are that a real defect signal is selected, defect data points and detection parameters are extracted, other eddy current detection data without defect signals are selected, and the defect data is enlarged or reduced according to the collection speed. Through this step, the length of the defect signal affected by the difference in probe speed during different data collection can be solved, and the length of the fused defect can be kept unchanged. The processed data is fused into the eddy current detection data without defect signals, mainly including calculating the deviation of the defect signal and the centering signal, and scaling the deviation by a normalized coefficient ratio, realizing the consistency of the fused data and the defect data, and guaranteeing that the new defect signal is generated according to the real one. DETAILED DESCRIPTION

[0049] The present application will be further described in detail below in combination with specific embodiments.

[0050] A steam generator heat pipe eddy current detection defect signal generation method selects signal eddy current detection data containing defects, extracts defect signal original data points and defect calibration related parameters, selects other eddy current detection data as target data, fuses the defect signal into a specified position of the target data, and forms a new defect signal. The specific steps are as follows:

[0051] Step 1: Defect signal information extraction

[0052] Step 11: Open the data group containing the defect signal, calibrate the signal according to the calibration tube data, and obtain related calibration parameters: including the rotation angle value of each channel, marked as Rotation; the voltage normalization coefficient Voltscale of each channel and the center point NullPoint (eddy current signal data point, including two components) of the heat pipe without defects set by setting;

[0053] Step 12: read the eddy current detection data of the defect signal, display the signal by using a bar graph and a Lisa chart, select the defect signal, determine the starting point of the defect signal, A point (the starting point) and B point (the ending point), and all data points between A point and B point as the data points PointNum1 of the defect signal, extract the data point set in the current channel starting point range to form the defect data point set of the current channel;

[0054] Step 13: calculate the defect tube probe movement speed Speed1, the calculation method is: the total length of the defect tube divided by the ratio of the effective data point number of the defect tube eddy current signal in the tube to the sampling rate;

[0055] Step 14: calculate the defect length, the calculation method is: the data point number of the defect starting point divided by the sampling rate, and then multiplied by the speed Speed1 to obtain the defect length Len1.

[0056] Step 2: align the defect signal data point set based on the selected other eddy current detection data

[0057] Step 21: refer to step 1, obtain and calculate the related parameters of the data group where the other eddy current detection data is located, including the sampling rate Sample2 and the probe movement speed Speed2;

[0058] Step 22: calculate the data point number of the other eddy current detection data corresponding to the defect length Len1, the calculation method is: the defect length Len1 divided by the speed Speed2 and then multiplied by the Sample2 to obtain the data point number PointNum2;

[0059] Step 23: when PointNum2 is less than PointNum1, compress the defect data point set to the length equal to PointNum1, when PointNum2 is greater than PointNum1, expand the defect data point set to the length equal to PointNum1, and when they are equal, do not process;

[0060] Step 24: the compression case, the processing process is as follows:

[0061] Calculate the compression coefficient PointNum1 / PointNum2;

[0062] Create an array DataNew[PointNum2] to save data points for the data segment to be reduced, each data unit being the eddy current signal data point (including X and Y components) for the PointNum1 data points Data[PointNum1] of the data segment, from the first point to the last point, sequentially perform the reduction processing, assuming that the current data point index is I, the corresponding newly generated data index is J, J = (INT) (I / (PointNum1 / PointNum2)), the data point is DataNew[J];

[0063] Using the Vmax measurement method, calculate the amplitude of Data[I] and DataNew[J] to the center point respectively, assuming that the amplitude calculated by Data[I] is Vi, and the amplitude calculated by DataNew[J] is Vj, compare Vi and Vj, if Vi is greater than Vj, then execute the following assignment DataNew[J] = Data[I]; until m data points are all converted, DataNew array is used as the compressed data.

[0064] In the expansion case, the processing algorithm is as follows:

[0065] Calculate the expansion coefficient PointNum2 / PointNum1;

[0066] For the PointNum1 data points Data[PointNum1], from the first point to the last point, sequentially calculate the data points that need to be interpolated between adjacent two data points, and use the linear difference method to calculate and interpolate the X and Y components of each data point. Form the data set of DataNew[PointNum2] as the data point set of the defect signal.

[0067] Step 3: Calculate the offset of all data points of the defect signal from the center point

[0068] Step 31: Calculate the center point phase calibration data point refNull, the calculation method is:

[0069] Refer to the rotation formula in the plane rectangular coordinate system, set the rotation angle of the coordinate axis as θ, P is any point on the plane, the coordinates in the original coordinate system xOy are (x, y), and the coordinates in the new coordinate system x'Oy' are (x', y'), then

[0070]

[0071] Through the above formula, rotate the reference channel center point by the rotation angle Rotation to obtain a new data point, marked as refNull;

[0072] Step 32: Calculate the offset of the defect signal data point set relative to the reference center point, forming a deviation value set with the same number of defect signal data points, the calculation method is as follows:

[0073] Rotate each data point in the defect signal data point set by an angle of Rotation, the calculation method is as follows: refer to the first step of step 3, and obtain the rotated data point set.

[0074] Calculate the deviation of each data point in the data point set, the calculation method is as follows: subtract the x of refNull point from the x of each point, and subtract the y of refNull point from the y of each point, to obtain the deviation point. Save all deviation points to the deviation value data set and mark it as PointList.

[0075] Step 4: Fusion of other target eddy current data

[0076] Step 41: Open the eddy current detection data set containing target data, calibrate the signal according to the calibration tube data, and obtain the relevant calibration parameters: including the rotation angle value of each channel, marked as Rotation2; the voltage normalization coefficient Voltscale2 of each channel and the center point NullPoint2 by setting the defect-free position in the heat pipe as the channel;

[0077] Step 42: Select the fusion data region of the target data, including the data start point (start2) and the data end point (end2), and the data length is PointNum2;

[0078] Step 43: Calculate the voltage normalization coefficient ratio of the target data channel (TChannel) and the defect data channel (Channel), marked as belt1:

[0079] Step 44: Fusion processing of the data points in the current target data channel fusion data region:

[0080] The target data channel fusion data region ranges from all data points with index start2 to end2, and the current data point is marked as CurDP, the index is index, and the corresponding defect deviation data set PointList index is index-start2. The two points are fused. The steps are as follows:

[0081] Calculate the data point after rotating the current data point by Rotation2, and mark it as Base;

[0082] The data point with PointList index index-start2 is marked as Crack; calculate x1 equal to the X of Crack divided by belt1, and y1 equal to the y of Crack divided by belt1;

[0083] adding x of CurDP to x1 and y of CurDP to y1 to obtain a data point NextDP;

[0084] rotating the data point obtained by rotating NextDP by Rotation2*(-1) degrees as a newly generated data point after fusing CurDP.

[0085] saving all the fused data points to the retList object as new data after fusing the defect data of the target channel, and the fusion information contained in the data is the newly generated defect signal.

Claims

1. A method for generating defect signals for eddy current detection in heat transfer tubes of a steam generator, characterized in that: Select eddy current detection data containing defects, extract the original data points of the defect signal and relevant parameters for defect calibration, select other eddy current detection data as target data, and fuse the defect signal into a specified position of the target data to form a new defect signal. The specific steps are as follows: Step 1: Defect signal information extraction; Step 2: Align the defect signal data point set based on other eddy current detection data; Step 3: Calculate the offset of all data points of the defect signal from the center point; Step 4: Fuse eddy current data from other targets.

2. The method for generating defect signals for eddy current detection of heat transfer tubes in a steam generator as described in claim 1, characterized in that, Step 1 includes: Step 11: Open the data group containing defective signals, calibrate the signals according to the calibration tube data, and obtain relevant calibration parameters: including the rotation angle value of each channel, marked Rotation; the voltage normalization coefficient Voltscale of each channel, and the eddy current signal data points, including two components, by setting the defect-free point in the heat transfer tube as the center point of the channel NullPoint; Step 12: Read the eddy current detection data containing the defect signal, display the signal using a bar chart or Lissajous figure, select the defect signal, determine the starting point of the defect signal, and take all data points between point A as the starting point and point B as the ending point as the data point PointNum1 of the defect signal. Extract the set of data points within the range of the starting point of the current channel to form the defect data point set of the current channel. Step 13: Calculate the speed Speed1 of the defect tube probe. The calculation method is: the total length of the defect tube divided by the ratio of the number of effective data points in the defect tube eddy current signal tube to the sampling rate. Step 14: Calculate the defect length. The calculation method is to divide the number of data points at the defect starting point by the sampling rate, and then multiply it by the speed Speed1 to obtain the defect length Len1.

3. The method for generating defect signals for eddy current detection in heat transfer tubes of a steam generator as described in claim 1, characterized in that, Step 2 includes: Step 21: Obtain and calculate the relevant parameters of the data group containing other eddy current detection data, including sampling rate Sample2 and probe movement speed Speed2; Step 22: Calculate the number of data points of other eddy current detection data corresponding to the defect length Len1. The calculation method is: divide the defect length Len1 by the speed Speed2 and then multiply by Sample2 to get the number of data points PointNum2. Step 23: When PointNum2 is less than PointNum1, compress the defect data point set to a length equal to PointNum1. When PointNum2 is greater than PointNum1, expand the defect data point set to a length equal to PointNum1. When they are equal, do not process them. Step 24: Process different compression scenarios.

4. The method for generating defect signals for eddy current detection in heat transfer tubes of a steam generator as described in claim 3, characterized in that, Step 24 includes: Calculate the compression ratio PointNum1 / PointNum2; Create an array DataNew[PointNum2] to store data points for this data segment to be reduced. Each data unit is an eddy current signal data point including two components, X and Y. For the data segment PointNum1 data points Data[PointNum1], perform reduction processing sequentially from the first point to the last point. Assume that the current data point index is I, and its corresponding newly generated data index is J, J = (INT)(I / (PointNum1 / PointNum2)), and the data point is DataNew[J]. Using the Vmax measurement method, the amplitudes of Data[I] and DataNew[J] to the center point are calculated respectively. Assuming that the amplitude calculated for Data[I] is Vi and the amplitude calculated for DataNew[J] is Vj, Vi and Vj are compared. If Vi is greater than Vj, the following assignment is performed: DataNew[J] = Data[I]; until all m data points have been converted, the DataNew array is used as the compressed data.

5. The method for generating defect signals for eddy current detection in heat transfer tubes of a steam generator as described in claim 3, characterized in that, Step 24 includes the following processing for the expanded case: Calculate the scaling factor PointNum2 / PointNum1; For the data points Data[PointNum1] of the data segment PointNum1, from the first point to the last point, the data points that need to be interpolated between adjacent data points are calculated in turn. The linear interpolation method is used to calculate and interpolate the X and Y components of each data point to form the dataset DataNew[PointNum2] as the data point set of the defect signal.

6. The method for generating defect signals for eddy current detection in heat transfer tubes of a steam generator as described in claim 1, characterized in that, Step 3 includes: Step 31: Calculate the refNull of the data points after center point phase calibration. Calculation method: Using the rotation formula in a Cartesian coordinate system, let the rotation angle of the coordinate axis be θ, and let P be any point in the plane with coordinates (x, y) in the original coordinate system xOy and (x′, y′) in the new coordinate system x′Oy′. Using the above formula, rotate the center point of the reference channel by the rotation angle Rotation to obtain a new data point, which is marked as refNull; Step 32: Calculate the offset of the defect signal data point set relative to the reference center point, forming a set of deviation values ​​that is the same as the number of defect signal data points.

7. The method for generating defect signals for eddy current detection in heat transfer tubes of a steam generator as described in claim 6, characterized in that, The calculation method for step 32 is as follows: Each data point in the defect signal data point set is rotated by a rotation angle of Rotation, and the rotated data point set is calculated. For each data point in the data point set, the deviation is calculated by subtracting the x of the refNull point from the x of each point and subtracting the y of the refNull point from the y of each point. The deviation points are then saved to the deviation value dataset and labeled as PointList.

8. The method for generating defect signals for eddy current detection of heat transfer tubes in a steam generator as described in claim 1, characterized in that, Step 4 includes: Step 41: Open the eddy current detection data group containing the target data, calibrate the signal according to the calibration tube data, and obtain the relevant calibration parameters: including the rotation angle value of each channel, marked Rotation2; the voltage normalization coefficient Voltscale2 of each channel, and NullPoint2, which is set as the center point of the channel by setting a defect-free point inside the heat transfer tube. Step 42: Select the fusion data region of the target data, including the data start point and data end point, with a data length of PointNum2; Step 43: Calculate the ratio of the voltage normalization coefficients of the target data channel and the defective data channel, denoted as belt1: Step 44: Perform fusion processing on the data points within the current target data channel fusion data area.

9. The method for generating defect signals for eddy current detection in heat transfer tubes of a steam generator as described in claim 5, characterized in that, The fusion process in step 44 includes fusing all data points in the target data channel within the range of index start2 to end2. Let the current data point be identified as CurDP and its index be index. The corresponding defect deviation dataset PointList index is index-start2. The two points are then fused.

10. The method for generating defect signals for eddy current detection of heat transfer tubes in a steam generator as described in claim 9, characterized in that, The fusion process in step 44 includes the following: After calculating Rotation2 after rotating the current data point, the data point is obtained and identified as Base; The data point with index index-start2 in PointList is identified as Crack; x1 is calculated as Crack's X divided by belt1, and y1 is calculated as Crack's y divided by belt1; Add x1 to x in CurDP and add y1 to y in CurDP to get the data point NextDP; The data points obtained by rotating NextDP by Rotation2*(-1) degrees are used as the newly generated data points after CurDP fusion. All fused data points are saved to a retList object as new data after fusing defect data in the target channel. The fusion information contained in this data is the newly generated defect signal.

Images