Steel belt fault detection method for spindle conveying

Abstract

The invention relates to the technical field of fault detection, in particular to a steel belt fault detection method for spindle conveying, and the method comprises the steps: collecting echo signals received after an ultrasonic probe transmits ultrasonic waves to a steel belt in an operation process; obtaining each return band; calculating the echo time difference of each echo band; acquiring a peak-to-peak delay coefficient of each echo band; calculating an echo consistency coefficient of each echo band; obtaining all first quartered return bands; acquiring a steel belt echo attenuation value of the echo signal; after wavelet decomposition is carried out on the echo signal, a defect noise estimated value of each wavelet coefficient is obtained; and further obtaining a self-adaptive wavelet threshold value of each wavelet coefficient, and carrying out fault detection on the spindle transportation steel belt. According to the invention, the fault detection precision of the ultrasonic transmission mode is improved.

Description

Technical Field

[0001] The present application relates to the technical field of fault detection, and in particular to a method for detecting steel belt faults used for ingot conveying. Background Art

[0002] In the textile industry, spindles are key components of spinning machines, responsible for evenly stretching and twisting roving into fine yarn. The conveying efficiency and stability of the spindles directly affect the operating efficiency and product quality of the entire textile production line. Therefore, the steel belts responsible for transporting the spindles need to be wear-resistant, high-temperature resistant, and strong in tensile strength. The steel belts may be worn, broken, corroded, and other faults. If these faults are not detected and handled in time, they will not only lead to a decrease in production efficiency, but also may cause safety accidents, equipment damage, and personal injury.

[0003] The ultrasonic transmission method can penetrate deep into the steel strip material, so the use of ultrasound to detect steel strip faults can detect microscopic defects that cannot be directly observed by the naked eye, and can be performed without stopping the machine, which is crucial to maintaining the stability of the steel strip and extending its service life. The noise generated by the spindle during operation will mix with the echo generated by ultrasonic testing of the steel strip, affecting the accuracy of the fault detection results. Summary of the invention

[0004] In order to solve the above technical problems, the present application provides a steel belt fault detection method for ingot conveying to solve the existing problems.

[0005] A steel belt fault detection method for ingot conveying in the present application adopts the following technical solution: An embodiment of the present application provides a method for detecting a steel belt fault for ingot conveying, the method comprising the following steps: Collect the echo signal received after the ultrasonic probe emits ultrasonic waves to the steel strip in the running process; obtain each echo segment in the echo signal; Obtain the echo time difference of each echo band according to the time difference between the peak values ​​of the echo bands; obtain the peak-to-peak delay coefficient of each echo band based on the difference between the echo time difference of each echo band and the average of the echo time difference; obtain the echo consistency coefficient of each echo band based on the distance difference between the echo bands and the peak-to-peak delay coefficient; Based on the first quartile of all echo consistency coefficients, all first quartile echo bands are obtained; based on the comparison between the peak value of each first quartile echo band and the peak value of the first echo band in the echo signal, the steel strip echo attenuation value of the echo signal is obtained; After performing wavelet decomposition on the echo signal, the amplitude distribution difference of the decomposed signal corresponding to each wavelet coefficient is obtained, and the defect noise estimation of each wavelet coefficient is obtained; Based on the attenuation value of the steel strip echo, the average value of the amplitude of each wavelet coefficient and the defect noise estimation, the adaptive wavelet threshold of each wavelet coefficient is obtained; according to the adaptive wavelet threshold, the denoised echo signal is obtained to perform fault detection on the ingot transport steel strip.

[0006] Further, the acquiring each echo segment in the echo signal includes: Extract the envelope of the echo signal, take the signal between the preset amplitude thresholds twice in a row and the signal whose amplitude average value is greater than the preset amplitude threshold as an echo segment, and obtain each echo segment in the echo signal.

[0007] Further, the method for obtaining the echo time difference is: The peak value of each echo band is obtained by using a peak detection algorithm, and the time difference between each echo band and the peak of its next adjacent echo is calculated as the echo time difference of each echo band.

[0008] Furthermore, the peak-to-peak delay coefficient is obtained by: ; In the formula, represents the peak-to-peak delay coefficient of the i-th echo segment, N is the total number of echo segments in the echo signal, represents the echo time difference of the i-th echo segment, is the average of the echo time differences of all echoes.

[0009] Further, the method for obtaining the echo consistency coefficient is: The average value of the distances between each echo band and all other echoes is calculated, and the ratio of the peak-to-peak delay coefficient of each echo band to the average value is taken as the echo consistency coefficient of each echo band.

[0010] Furthermore, the method for acquiring the first quarter echo band is: The echo consistency coefficients of all echoes are sorted from small to large, the first quartile of the echo consistency coefficient is obtained, and the echoes with echo consistency coefficients greater than the first quartile are taken as the first quartile echo segment.

[0011] Furthermore, the formula for the steel belt echo attenuation value is: Wherein, J is the steel strip echo attenuation value of the echo signal, W represents the total number of the first quarter echo band, ln() represents the logarithmic function with the natural constant e as the base, represents the peak amplitude of the jth first quarter echo band, Indicates the peak amplitude of the first echo in the echo signal.

[0012] Furthermore, the defect noise estimation is obtained by: Perform wavelet decomposition on the echo signal to obtain each wavelet coefficient; The difference between the median of the amplitude of the decomposed signal corresponding to each wavelet coefficient and the mean absolute deviation is calculated as the defect noise estimation of each wavelet coefficient.

[0013] Furthermore, the method for obtaining the adaptive wavelet threshold is: ; In the formula, is the adaptive wavelet threshold of the mth wavelet coefficient, is the amplitude sequence of the decomposed signal corresponding to the mth wavelet coefficient, J is the attenuation value of the steel strip echo, Represents the defect noise estimation of the mth wavelet coefficient, avg() is the average function, and norm() is the normalization function.

[0014] Further, the acquisition of the denoised echo signal and the detection of faults of the ingot transport steel strip include: De-noising the wavelet coefficients according to the adaptive wavelet threshold of each wavelet coefficient, reconstructing all the de-noised wavelet coefficients to obtain the echo signal with the spindle noise removed; If, in addition to the front echo and the rear echo, the echo signal of the steel strip to be detected contains an echo whose amplitude is greater than the preset amplitude threshold, or an echo whose width exceeds one-fourth of the front echo, then the steel strip to be detected is faulty; otherwise, the steel strip is not faulty.

[0015] This application has at least the following beneficial effects: The present application collects the echo signal of the steel strip during the operation of the ingot conveying, and divides the echo signal into echo bands, so as to more accurately analyze the noise of each part of the echo signal, and obtain the peak-to-peak delay coefficient according to the difference between the peak values ​​of the echo bands and the average of the differences, which reflects the consistency of the intervals between the echo bands, and then reflects the geometric characteristics inside the steel strip. The echo consistency coefficient is obtained according to the peak-to-peak delay coefficient, so as to more accurately analyze whether there are internal defects in the steel strip, and then construct the steel strip echo attenuation value, which better reflects the attenuation of the sound wave in the steel strip, and then calculate the defect noise estimation to estimate the noise of the defect echo, analyze the relationship between the internal defects and noise of the steel strip according to the estimation, and obtain the adaptive wavelet threshold of each wavelet coefficient; finally, denoising is completed to obtain the reconstructed echo signal, and the result of steel strip fault detection is obtained according to the echo signal with the spindle noise removed.

[0016] Its beneficial effect is that a threshold is obtained according to the attenuation characteristics of the echo signal when there is an internal defect in the steel strip combined with the characteristics of the spindle noise, thereby using the echo signal to detect internal faults in the steel strip, and then accurately detecting the steel strip fault while filtering out noise interference, thereby improving the fault detection accuracy of the ultrasonic transmission method. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] In order to more clearly illustrate the technical solutions and advantages in the embodiments of the present application or the prior art, the drawings required for use in the embodiments or the prior art descriptions are briefly introduced below. Obviously, the drawings described below are only some embodiments of the present application. For ordinary technicians in this field, other drawings can be obtained based on these drawings without creative work.

[0018] Figure 1 A flow chart of a steel belt fault detection method for ingot conveying provided in the present application; Figure 2 A flowchart of adaptive wavelet threshold acquisition provided for one embodiment of the present application. DETAILED DESCRIPTION

[0019] In order to further explain the technical means and effects adopted by the present application to achieve the predetermined invention purpose, the following is a detailed description of the steel belt fault detection method for spindle conveying proposed in the present application, its specific implementation, structure, features and effects, in combination with the accompanying drawings and preferred embodiments. In the following description, different "one embodiment" or "another embodiment" does not necessarily refer to the same embodiment. In addition, specific features, structures or characteristics in one or more embodiments may be combined in any suitable form.

[0020] Unless defined otherwise, all technical and scientific terms used herein have the same meaning as commonly understood by one of ordinary skill in the art to which this application belongs.

[0021] The specific scheme of a steel belt fault detection method for ingot conveying provided by the present application is described in detail below with reference to the accompanying drawings.

[0022] An embodiment of the present application provides a method for detecting a steel belt fault for ingot conveying. Specifically, the following method for detecting a steel belt fault for ingot conveying is provided. Figure 1 , the method comprises the following steps: Step S1, collecting the echo signal received after the ultrasonic probe emits ultrasonic waves to the steel strip in the running process; obtaining each echo segment in the echo signal.

[0023] An ultrasonic probe is used to transmit ultrasonic waves on a running steel belt used for ingot conveying. The ultrasonic frequency used for detection is 2.5 MHz, the wafer size of the ultrasonic probe is 20 mm, the probe type is a single crystal straight probe, and the probe coupling agent is alcohol.

[0024] The receiver is used to receive the echo signal, and the echo signal received by the receiver is amplified to improve the sensitivity of detection so that even weak signals can be detected. Signal amplification and ultrasonic flaw detection are well-known technologies, and the specific process will not be repeated.

[0025] Step S2, obtaining the echo time difference of each echo band according to the time difference between the peak values ​​of the echo bands; obtaining the peak-to-peak delay coefficient of each echo band based on the difference between the echo time difference of each echo band and the average of the echo time difference; obtaining the echo consistency coefficient of each echo band based on the distance difference between the echo bands and the peak-to-peak delay coefficient.

[0026] When the steel belt is running for a long time in the task of conveying ingots, fatigue, wear or other forms of failure may occur. In order to ensure production efficiency and safety, ultrasonic waves can be used to detect locations prone to internal defects to evaluate the integrity of its structure.

[0027] In the actual detection process, the ultrasonic probe emits high-frequency sound waves to penetrate the steel strip. When encountering internal defects, the sound waves will be reflected and return to the probe, and the receiver will receive the echo signal. However, these returned signals are often mixed with the noise generated by the operation of the spindle. If not properly processed, these noises will seriously affect the accuracy of the detection results.

[0028] Furthermore, this embodiment uses wavelet transform to denoise the echo signal. The threshold in wavelet transform is a parameter used for denoising, which determines which coefficients in the wavelet domain are considered as noise and need to be suppressed or removed, and which coefficients are considered as part of the signal and need to be retained.

[0029] In the wavelet transform denoising process, the setting of the threshold is crucial to the denoising effect. In order to obtain a better denoising effect, this embodiment determines the threshold according to the characteristics of the noise generated by the spindle operation and the coefficient characteristics after the wavelet transform.

[0030] Specifically, when the ultrasonic probe emits an ultrasonic pulse into the steel strip, the sound waves propagate inside the material. If the sound waves encounter internal structural changes in the material during propagation, such as grain boundaries, inclusions, cracks or other defects, they will be reflected at these interfaces. These reflected sound waves then return to the probe, forming echo signals that can be captured by the detection system.

[0031] The echo signal is usually displayed as a waveform on an oscilloscope or detection software. The horizontal axis represents time, while the vertical axis represents amplitude. A healthy steel strip only shows simple echoes from both sides of the material surface. If there are internal defects, additional echoes will be displayed at the corresponding time or distance position.

[0032] Specifically, when the ultrasonic pulse is emitted from the probe and encounters the surface of the steel strip for the first time, an initial echo is generated, called the front echo. The amplitude of the front echo is usually higher because it is the direct reflection of the ultrasonic wave when it first contacts the material. The ultrasonic wave penetrates the steel strip and generates a second main echo when it reaches the other side surface, called the back echo.

[0033] When ultrasonic waves propagate inside the steel strip and encounter defects, they are reflected at these defect interfaces to form defect echoes. The time and amplitude of the defect echo depend on the type, size, shape and position of the defect relative to the ultrasonic beam. The defect echo is located between the front and rear echoes in time, and the amplitude may vary depending on the nature of the defect. It is usually smaller than the surface echo but sufficient to be identified.

[0034] Based on the above analysis, the amplitude of ultrasonic waves will gradually decrease during the propagation of the steel strip, that is, attenuation occurs. There is a certain relationship between the attenuation degree of ultrasonic waves and the wavelet threshold. Therefore, the initial threshold can be set by first calculating the steel strip echo attenuation value according to the attenuation degree of ultrasonic waves in the steel strip.

[0035] When using ultrasonic waves to detect the steel belts that convey spindles, non-stop testing is often performed to ensure production efficiency. At this time, the noise generated by the operation of the spindles will mix with the echo, causing the attenuation characteristics of the echo to be masked or distorted, which is not conducive to signal interpretation.

[0036] The attenuation between different echoes may be destroyed. It is necessary to exclude echoes with different attenuation from other echoes to avoid large errors in the obtained echo attenuation coefficients, which will affect the calculation of the wavelet threshold. Therefore, it is necessary to first calculate the consistency between different echoes and exclude echoes with low consistency.

[0037] In order to determine the difference between the echo generated by the surface reflection of the steel strip and the echo generated by the internal defects and other parts in the echo signal, it can be determined based on the envelope of the echo signal.

[0038] Specifically, this embodiment uses the moving average method to obtain the envelope of the echo signal, wherein the methods for extracting the envelope include Hilbert transform, moving average method, etc., and the implementer may select other methods for extracting the envelope according to the actual situation; it should be noted that the envelope extraction process is a well-known technology and will not be described in detail here. The signal between the amplitude thresholds that the envelope passes through twice in a row and the amplitude average value is greater than the amplitude threshold is regarded as an echo segment, and the value of the amplitude threshold in this embodiment is 0.1.

[0039] The peak value of each echo band is obtained by using the peak detection algorithm, and the time difference between each echo band and the peak of the next adjacent echo band is calculated as the echo time difference of each echo band. In this embodiment, the AMPD algorithm is used to implement peak detection, and the implementer can use other peak detection algorithms according to actual conditions. It should be noted that the last echo band is no longer calculated.

[0040] Furthermore, in order to reflect the geometric characteristics in the steel strip, the present application calculates the peak-to-peak delay coefficient according to the difference between the echo time difference of each echo segment and the average of the echo time difference, and the formula is: ; In the formula, represents the peak-to-peak delay coefficient of the i-th echo segment; N represents the total number of echo segments in the echo signal, represents the echo time difference of the i-th echo segment, is the average of the echo time differences of all echoes.

[0041] It should be noted that the peak-to-peak delay coefficient indicates the time interval between two echo bands. If the interior of the steel strip is a layered structure or has repetitive geometric features, or there are no defects or the defects are small, the time interval of the echo bands will be relatively consistent. If it is interfered by spindle noise, even if there are no defects, the time interval will be reduced, affecting the judgment of the internal defects of the steel strip.

[0042] Furthermore, the echo consistency coefficient of each echo band is obtained based on the peak-to-peak delay coefficient, and the formula is: ; In the formula, Represents the echo consistency coefficient of the i-th echo band; the i-th echo band represents the peak-to-peak delay coefficient of the i-th echo band, It represents the average value of the distance between the i-th echo band and all other echo bands. In this embodiment, the DTW distance is selected as the distance measurement method. The implementer can select other measurement distances according to the actual situation.

[0043] DTW distance indicates the correlation between echo bands. Since the echoes generated by the collision between ultrasonic waves and the surface or inside of the steel strip have a high correlation, the correlation between the echo band affected by spindle noise and other echo bands will decrease, making it easy to be identified as an internal defect.

[0044] At this point, the larger the echo consistency coefficient of each echo band, the higher the consistency between the echo band and other echo bands, the smaller the influence of spindle noise, and the greater the possibility of internal defects.

[0045] Step S3, obtaining all first quartile echo bands based on the first quartile of all echo consistency coefficients; obtaining the steel strip echo attenuation value of the echo signal based on the comparison between the peak values ​​of each first quartile echo band and the peak value of the first echo band in the echo signal.

[0046] The echo consistency coefficients of all echo bands are sorted from small to large, the first quartile of the echo consistency coefficient is obtained, and the echo bands with echo consistency coefficients greater than the first quartile are taken as the first quartile echo bands.

[0047] Furthermore, in order to reflect the attenuation of the sound wave in the steel strip, the present application calculates the steel strip echo attenuation value J based on the comparison of the peak value of each first quarter echo band and the peak value of the first echo band in the echo signal, and the formula is: ; Wherein, J represents the steel strip echo attenuation value of the echo signal, W represents the total number of the first quarter echo band, ln() represents the logarithmic function with the natural constant e as the base, represents the amplitude peak value of the j first quarter echo bands, Indicates the peak amplitude of the first echo segment in the echo signal.

[0048] The numerical value of the steel strip echo attenuation value indicates the attenuation degree of the ultrasonic wave in the steel strip. The larger the value, the more serious the attenuation of the ultrasonic wave during the propagation in the steel strip. Therefore, the amplitude of the rear echo may become very small due to attenuation. If the wavelet threshold is set too high, the real signal with a smaller amplitude due to attenuation may be mistakenly regarded as noise removal, resulting in missed detection.

[0049] Therefore, the greater the steel strip echo attenuation value, the smaller the wavelet threshold should be set, and vice versa. The steel strip echo attenuation value first excludes the echoes that are greatly interfered by the spindle noise in the original echo signal, taking into account the characteristics of the spindle noise and the ultrasonic attenuation characteristics, so the obtained wavelet threshold can better remove the spindle noise.

[0050] Step S4, after performing wavelet decomposition on the echo signal, the amplitude distribution difference of the decomposed signal corresponding to each wavelet coefficient is obtained, and the defect noise estimation of each wavelet coefficient is obtained.

[0051] In order to better identify the echo signals generated by defects, the wavelet transform is used to process the echo signals. First, the echo signals should be decomposed by wavelet, and then an adaptive wavelet threshold is set for the wavelet coefficients of each decomposed signal obtained by the decomposition. The wavelet threshold is used to distinguish the details of the noise signal from the source signal. The wavelet coefficients less than the wavelet threshold are generated by noise, and the wavelet coefficients greater than or equal to the wavelet threshold are generated by the source signal. The denoised wavelet coefficients are then reconstructed to obtain the denoised echo signal.

[0052] In this embodiment, the wavelet function selected for wavelet transform of the echo signal is the Daubechies wavelet function, the N value is 3, and the echo signal is decomposed into 4 layers to obtain various wavelet coefficients.

[0053] Since there is spindle noise in the echo signal, and the noise is decomposed into different wavelet coefficients after wavelet transform, and the spindle noise will affect the echo signal generated by the defect, in order to consider the different characteristics between the wavelet coefficients obtained after wavelet transform, the steel strip defect noise level is estimated according to different wavelet coefficients to obtain the adaptive wavelet threshold. The wavelet coefficient includes real and imaginary parts, where the real part represents the amplitude of the signal at different scales and positions, that is, the amplitude, and the imaginary part represents the phase information.

[0054] The defect noise level of the steel strip can be obtained by the median and mean absolute deviation of the amplitude of the decomposed signal corresponding to the wavelet coefficient, because the median, as a robust statistic, is insensitive to outliers. In actual industrial environments, the steel strip may have grease, rust or other surface irregularities, which may cause outliers in the signal, but these outliers are not related to the internal defects of the steel strip and should be excluded. Using the median as the basis for noise estimation can reduce the impact of these outliers on noise estimation.

[0055] The mean absolute deviation (MAD) is a measure of the deviation of a data point from the central trend, which can adapt to the characteristics of different noise sources and provide a more comprehensive description for noise estimation. The mean absolute deviation is a well-known technology and will not be described in detail in this embodiment.

[0056] Therefore, the combined use of the median and mean absolute deviation helps to distinguish between normal signal changes caused by these microstructures and abnormal changes caused by noise, and calculates the defect noise estimate for each wavelet coefficient as follows: ; In the formula, represents the defect noise estimate of the mth wavelet coefficient; It represents the amplitude sequence of the decomposed signal corresponding to the mth wavelet coefficient, median() represents the calculation of the median, and mad() represents the calculation of the mean absolute deviation.

[0057] The numerical value of the defect noise estimation indicates the defect noise estimation level in the current wavelet coefficient. The larger its value is, the more spindle noise there is in the current wavelet coefficient, and the proportion of echoes caused by defects is relatively small. Therefore, a larger wavelet threshold should be set. On the contrary, a smaller wavelet threshold should be selected to remove the spindle noise more effectively.

[0058] Step S5, based on the steel strip echo attenuation value, the average value of the amplitude of each wavelet coefficient and the defect noise estimation, obtain the adaptive wavelet threshold of each wavelet coefficient; obtain the denoised echo signal according to the adaptive wavelet threshold, and perform fault detection on the ingot transport steel strip.

[0059] The initial wavelet threshold of each wavelet coefficient is set to the average value of the amplitude of each wavelet coefficient. The adaptive wavelet threshold of each wavelet coefficient is obtained by adjusting the attenuation value of the steel strip echo and the spindle noise estimation. The formula is: ; In the formula, represents the adaptive wavelet threshold of the mth wavelet coefficient, represents the amplitude sequence of the decomposed signal corresponding to the mth wavelet coefficient, J represents the attenuation value of the steel strip echo, represents the defect noise estimation of the mth wavelet coefficient, avg() represents the average function, and norm() represents the normalization function. The flowchart of adaptive wavelet threshold acquisition is as follows: Figure 2 shown.

[0060] In the formula, the adaptive wavelet coefficient is adjusted according to the steel strip echo attenuation value and the defect noise estimation. When the steel strip echo attenuation value of the echo signal is larger, the ultrasonic wave is more severely attenuated during the propagation in the steel strip, so the amplitude of the rear echo may become very small due to attenuation. At this time, the wavelet threshold should be reduced. When the defect noise estimation is small, the less noise in the wavelet coefficient, the smaller the echo proportion caused, and the wavelet threshold should be reduced; otherwise, the wavelet threshold should be increased.

[0061] At this point, the adaptive wavelet threshold of each wavelet coefficient is obtained.

[0062] The wavelet coefficients are denoised according to the adaptive wavelet threshold of each wavelet coefficient, and all the denoised wavelet coefficients are reconstructed to obtain the echo signal without spindle noise. Wavelet transform denoising and signal reconstruction belong to well-known technologies, and the specific process will not be repeated here.

[0063] The ultrasonic probe is evenly moved on the steel strip to be detected, and each echo band signal is processed according to the above steps to obtain the reconstructed echo band signals.

[0064] The internal defect fault of the steel strip used for ingot conveying is analyzed based on the processed echo signal. The first echo in the echo signal usually has a larger amplitude. This is the echo on the front surface of the steel strip. Similarly, the last echo has a smaller amplitude. This is the rear echo generated by the ultrasonic wave propagating inside the steel strip and colliding with the rear side of the steel strip after attenuation.

[0065] The amplitude of the echo is usually proportional to the size of the defect. A larger defect will generally produce a higher echo signal. The shape of the echo signal provides clues about the type of defect. For example, a sharp echo may indicate a small hole or crack, while a wide echo may indicate a larger area defect.

[0066] In addition, if there are multiple defects in the material, multiple echoes may be observed in the echo signal, each echo corresponding to a different defect or interface.

[0067] If, in addition to the front echo and the rear echo, the echo signal of the steel strip to be tested contains a defect echo whose echo amplitude is greater than a preset amplitude threshold, or an echo whose width exceeds one-fourth of the front echo, the steel strip to be tested is faulty; otherwise, the steel strip is not faulty.

[0068] So far, a steel belt fault detection method for ingot conveying has been completed.

[0069] It should be noted that the above-mentioned sequence of the embodiments of the present application is only for description and does not represent the advantages and disadvantages of the embodiments. The above-mentioned specific embodiments of the present application are described. In addition, the processes depicted in the accompanying drawings do not necessarily require the specific order or continuous order shown to achieve the desired results. In some embodiments, multitasking and parallel processing are also possible or may be advantageous.

[0070] The various embodiments in the present application are described in a progressive manner, and the same or similar parts between the various embodiments can be referenced to each other, and each embodiment focuses on the differences from other embodiments.

[0071] The embodiments described above are only used to illustrate the technical solutions of the present application, rather than to limit them. Modifications to the technical solutions recorded in the aforementioned embodiments, or equivalent replacement of some of the technical features therein, do not cause the essence of the corresponding technical solutions to deviate from the scope of the technical solutions of the embodiments of the present application, and should all be included in the protection scope of the present application.

Claims

1. A method for detecting steel belt faults for spindle conveying, characterized in that: The method comprises the following steps: Collect the echo signal received after the ultrasonic probe emits ultrasonic waves to the steel strip in the running process; obtain each echo segment in the echo signal; Obtain the echo time difference of each echo band according to the time difference between the peak values ​​of the echo bands; obtain the peak-to-peak delay coefficient of each echo band based on the difference between the echo time difference of each echo band and the average of the echo time difference; obtain the echo consistency coefficient of each echo band based on the distance difference between the echo bands and the peak-to-peak delay coefficient; Based on the first quartile of all echo consistency coefficients, all first quartile echo bands are obtained; based on the comparison between the peak value of each first quartile echo band and the peak value of the first echo band in the echo signal, the steel strip echo attenuation value of the echo signal is obtained; After performing wavelet decomposition on the echo signal, the amplitude distribution difference of the decomposed signal corresponding to each wavelet coefficient is obtained, and the defect noise estimation of each wavelet coefficient is obtained; Based on the attenuation value of the steel strip echo, the average value of the amplitude of each wavelet coefficient and the defect noise estimation, the adaptive wavelet threshold of each wavelet coefficient is obtained; according to the adaptive wavelet threshold, the denoised echo signal is obtained to perform fault detection on the ingot transport steel strip.

2. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The step of acquiring each echo segment in the echo signal comprises: Extract the envelope of the echo signal, take the signal between the preset amplitude thresholds twice in a row and the signal whose amplitude average value is greater than the preset amplitude threshold as an echo segment, and obtain each echo segment in the echo signal.

3. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The method for obtaining the echo time difference is: The peak value of each echo band is obtained by using a peak detection algorithm, and the time difference between each echo band and the peak of its next adjacent echo is calculated as the echo time difference of each echo band.

4. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The method for obtaining the peak-to-peak delay coefficient is: ; In the formula, represents the peak-to-peak delay coefficient of the i-th echo segment, N is the total number of echo segments in the echo signal, represents the echo time difference of the i-th echo segment, is the average of the echo time differences of all echoes.

5. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The method for obtaining the echo consistency coefficient is: The average value of the distances between each echo band and all other echoes is calculated, and the ratio of the peak-to-peak delay coefficient of each echo band to the average value is taken as the echo consistency coefficient of each echo band.

6. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The method for obtaining the first quarter echo band is: The echo consistency coefficients of all echoes are sorted from small to large, the first quartile of the echo consistency coefficient is obtained, and the echoes with echo consistency coefficients greater than the first quartile are taken as the first quartile echo segment.

7. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The formula for the steel belt echo attenuation value is: Wherein, J is the steel strip echo attenuation value of the echo signal, W represents the total number of the first quarter echo band, ln() represents the logarithmic function with the natural constant e as the base, represents the peak amplitude of the jth first quarter echo band, Indicates the peak amplitude of the first echo in the echo signal.

8. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The method for obtaining the defect noise estimation is: Perform wavelet decomposition on the echo signal to obtain each wavelet coefficient; The difference between the median of the amplitude of the decomposed signal corresponding to each wavelet coefficient and the mean absolute deviation is calculated as the defect noise estimation of each wavelet coefficient.

9. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The method for obtaining the adaptive wavelet threshold is: ; In the formula, is the adaptive wavelet threshold of the mth wavelet coefficient, is the amplitude sequence of the decomposed signal corresponding to the mth wavelet coefficient, J is the attenuation value of the steel strip echo, Represents the defect noise estimation of the mth wavelet coefficient, avg() is the average function, and norm() is the normalization function.

10. A method for detecting steel belt faults for spindle conveying according to claim 1, characterized in that: The method of obtaining the denoised echo signal and performing fault detection on the ingot transport steel belt comprises: De-noising the wavelet coefficients according to the adaptive wavelet threshold of each wavelet coefficient, reconstructing all the de-noised wavelet coefficients to obtain the echo signal with the spindle noise removed; If, in addition to the front echo and the rear echo, the echo signal of the steel strip to be detected contains an echo whose amplitude is greater than the preset amplitude threshold, or an echo whose width exceeds one-fourth of the front echo, then the steel strip to be detected is faulty; otherwise, the steel strip is not faulty.

Images