Ultrasonic echo defect signal identification method, device and equipment and storage medium

Abstract

The invention relates to an ultrasonic echo defect signal identification method and device, equipment and a storage medium, and the method comprises the steps: carrying out the preprocessing of an obtained ultrasonic echo signal of a steel plate, and obtaining an ultrasonic echo signal interval; smoothing each target ultrasonic echo signal in the ultrasonic echo signal interval to obtain a smooth ultrasonic echo signal, and determining all maximum value points and all minimum value points in the ultrasonic echo interval based on the smooth ultrasonic echo signal; based on all the maximum value points and all the minimum value points, defect characteristic parameters in the ultrasonic echo interval are extracted, and based on the Bayesian optimization technology, defect characteristic parameter optimal solutions corresponding to the defect characteristic parameters are determined; based on the defect characteristic parameters and the optimal solution of the defect characteristic parameters, defect signal identification is conducted on the ultrasonic echo signal interval, and the defect state of the steel plate is determined based on the defect signal identification result; compared with the prior art, the technical scheme of the invention can improve the accuracy of steel plate defect identification.

Description

Technical Field

[0001] This application relates to the technical field of steel plate flaw detection, and particularly to a method, device, equipment and storage medium for identifying ultrasonic echo defect signals. Background Art

[0002] In modern industry, the quality and structural integrity of steel plates are crucial. To ensure the quality of steel plates, ultrasonic technology is usually used to detect steel plates. The core of ultrasonic detection technology lies in utilizing the physical properties such as reflection and refraction of ultrasonic waves when propagating inside the steel plate, which can effectively detect and identify various defects inside the steel plate material, such as pores, cracks, inclusions, etc.

[0003] During the ultrasonic detection process of steel plates, the detection results are mainly presented in the form of ultrasonic echo signals. In the prior art, when processing ultrasonic echo signals, the ultrasonic echo signals are usually processed into a waveform curve to reflect the propagation and reflection of ultrasonic waves inside the steel plate. Subsequently, the waveform curve is intercepted to extract the part of the waveform curve that is considered to possibly represent a defect from the entire waveform curve. Then, the intercepted defect waveform curve is calculated to extract peak data, and these peak data are stored in a defect class array. Finally, the defect class array is further processed, mainly using the defect threshold judgment method and morphological analysis method to determine the defect type of the steel plate according to the set threshold and the analysis of the defect morphology.

[0004] However, the above existing steel plate defect recognition methods have many limitations and deficiencies. First, when using the defect threshold judgment method for defect recognition, its recognition accuracy highly depends on the artificially set threshold. This uncertainty of artificial setting will bring large subjective errors to defect recognition, resulting in misjudgment and missed judgment. Second, during the entire recognition process, there is a lack of effective means to suppress and process noise interference, resulting in noise signals seriously affecting the final recognition result, making the recognition result unstable and thus generating a high error rate. Third, the existing morphological analysis methods mainly analyze based on the morphological characteristics of defects. This method is difficult to accurately reflect the real situation when facing defects with complex or irregular shapes, unable to comprehensively capture and analyze various characteristics of defects, thus missing some defect information and limiting its defect recognition ability for different types of defects. Summary of the Invention

[0005] This application provides a method, device, equipment and storage medium for identifying ultrasonic echo defect signals to improve the accuracy of steel plate defect recognition.

[0006] In a first aspect, the present application provides a method for identifying ultrasonic echo defect signals, including: obtaining an ultrasonic echo signal of a steel plate, preprocessing the ultrasonic echo signal to obtain an ultrasonic echo signal interval; smoothing each target ultrasonic echo signal within the ultrasonic echo signal interval to obtain a corresponding smoothed ultrasonic echo signal, and based on the smoothed ultrasonic echo signal, determining all maximum points and all minimum points within the ultrasonic echo interval; extracting defect feature parameters within the ultrasonic echo interval based on all the maximum points and all the minimum points, and determining an optimal solution of the defect feature parameters corresponding to the defect feature parameters based on the Bayesian optimization technique; identifying defect signals for the ultrasonic echo signal interval based on the defect feature parameters and the optimal solution of the defect feature parameters, and determining the defect state of the steel plate based on the defect signal identification result.

[0007] In a possible implementation manner, obtaining an ultrasonic echo signal of a steel plate, preprocessing the ultrasonic echo signal to obtain an ultrasonic echo signal interval specifically includes: obtaining an ultrasonic echo signal of a steel plate, where the ultrasonic echo signal includes an echo signal and a bottom wave signal; respectively calculating an amplitude difference between the amplitude of each echo signal and the amplitude of the bottom wave signal in the ultrasonic echo signal, and when it is determined that the amplitude differences corresponding to a continuous plurality of echo signals are greater than a preset amplitude difference threshold, determining an ultrasonic echo signal interval based on the continuous plurality of echo signals, and respectively using the continuous plurality of echo signals as target ultrasonic echo signals within the ultrasonic echo signal interval.

[0008] In a possible implementation manner, smoothing each target ultrasonic echo signal within the ultrasonic echo signal interval to obtain a corresponding smoothed ultrasonic echo signal specifically includes: obtaining a sequence of echo signal points corresponding to each target ultrasonic echo signal within the ultrasonic echo signal interval, and assigning a corresponding weight value to each echo signal point in the sequence of echo signal points; performing a weighted summation process on the sequence of echo signal points based on the weight value to obtain a weighted summation value of the echo signal points, and performing a summation process on all the weight values to obtain a total weight; calculating a corresponding smoothed ultrasonic echo signal for each target ultrasonic echo signal within the ultrasonic echo signal interval based on the weighted summation value of the echo signal points and the total weight.

[0009] In a possible implementation manner, based on the smoothed ultrasonic echo signal, all the maximum points and all the minimum points within the ultrasonic echo interval are determined. Specifically, it includes: traversing each smoothed ultrasonic echo signal within the ultrasonic echo signal interval, and during the traversal process, obtaining the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal corresponding to the current smoothed ultrasonic echo signal; when the current smoothed ultrasonic echo signal is respectively greater than the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal, taking the current smoothed ultrasonic echo signal as a maximum point; when the current smoothed ultrasonic echo signal is respectively less than the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal, taking the current smoothed ultrasonic echo signal as a minimum point; until the traversal is completed, integrating to obtain all the maximum points and all the minimum points within the ultrasonic echo interval.

[0010] In a possible implementation manner, based on all the maximum points and all the minimum points, defect characteristic parameters within the ultrasonic echo interval are extracted. Specifically, it includes: respectively obtaining the adjacent minimum points corresponding to each maximum point, performing combination processing on each maximum point and the adjacent minimum points corresponding to each maximum point to obtain a plurality of extreme point pairs; respectively calculating the span values between the target maximum point and the target adjacent minimum point in each extreme point pair to obtain a two-dimensional span value sequence; obtaining the target number of first pre-ultrasonic echo signals corresponding to the target maximum value in each extreme point pair, and obtaining the target number of first post-ultrasonic echo signals corresponding to the target minimum value in each extreme point pair, respectively calculating the amplitude difference between the amplitude mean value of the first pre-ultrasonic echo signals corresponding to each extreme point pair and the amplitude mean value of the first post-ultrasonic echo signals to obtain a two-dimensional amplitude difference sequence; obtaining the first echo signal point sequence corresponding to each maximum point, respectively calculating the first data change rate of adjacent ultrasonic echo signals in the first echo signal sequence, and based on the first data change rate, determining the first data change rate range corresponding to each maximum point, and based on the first data change rate range, obtaining a first two-dimensional data change rate range sequence; obtaining the second echo signal point sequence corresponding to each maximum point, respectively calculating the second data change rate of adjacent ultrasonic echo signals in the second echo signal sequence, and based on the second data change rate, determining the second data change rate range corresponding to each maximum point, and based on the second data change rate range, obtaining a second two-dimensional data change rate range sequence; taking the two-dimensional span value sequence, the two-dimensional amplitude difference sequence, the target number, the first two-dimensional data change rate range sequence, and the second two-dimensional data change rate range sequence as the defect characteristic parameters within the ultrasonic echo interval.

[0011] In a possible implementation manner, based on the Bayesian optimization technique, the optimal solution of the defect feature parameters corresponding to the defect feature parameters is determined, specifically including: based on the defect feature parameters, an objective function is constructed, and based on the defect feature parameters and the objective function values corresponding to the defect feature parameters, a Gaussian process regression model is constructed; based on a hybrid acquisition strategy with dynamically adjusted weights, an acquisition function is used to determine the evaluated defect feature parameters, and based on the evaluated defect feature parameters, model iteration processing is performed on the Gaussian process regression model until the target defect feature parameters corresponding to the minimum value of the objective function are determined, and the target defect feature parameters are used as the optimal solution of the defect feature parameters corresponding to the defect feature parameters.

[0012] In a possible implementation manner, based on the defect feature parameters and the optimal solution of the defect feature parameters, defect signal identification is performed on the ultrasonic echo signal interval, specifically including: the defect feature parameters include a two-dimensional span value sequence, a two-dimensional amplitude difference sequence, the number of targets, a first two-dimensional data change rate range sequence, and a second two-dimensional data change rate range sequence; the optimal solution of the defect feature parameters includes an optimal span value, an optimal number of targets, an optimal amplitude difference determined based on the optimal number of targets, an optimal first data change rate range, and an optimal second data change rate range; when any target span value in the two-dimensional span value sequence is greater than the optimal span value, any target amplitude difference in the two-dimensional amplitude difference sequence is greater than the optimal amplitude difference, any target first data change rate range in the first two-dimensional data change rate range sequence belongs to the optimal first data change rate range, and any target second data change rate range in the second two-dimensional data change rate range sequence belongs to the optimal second data change rate range, it is determined that there is a defect signal in the ultrasonic echo signal interval; wherein, any target span value, any target amplitude difference, any target first data change rate range, and any target second data change rate range are calculated from the same extreme point pair.

[0013] Second aspect, the present application provides an ultrasonic echo defect signal recognition device, including: a signal preprocessing module, a signal smoothing processing module, a Bayesian optimization module, and a defect signal recognition module; wherein, the signal preprocessing module is configured to obtain an ultrasonic echo signal of a steel plate, perform preprocessing on the ultrasonic echo signal to obtain an ultrasonic echo signal interval; the signal smoothing processing module is configured to perform smoothing processing on each target ultrasonic echo signal within the ultrasonic echo signal interval to obtain a corresponding smoothed ultrasonic echo signal, and based on the smoothed ultrasonic echo signal, determine all maximum points and all minimum points within the ultrasonic echo interval; the Bayesian optimization module is configured to extract defect feature parameters within the ultrasonic echo interval based on all the maximum points and all the minimum points, and determine an optimal solution of the defect feature parameters corresponding to the defect feature parameters based on Bayesian optimization technology; the defect signal recognition module is configured to perform defect signal recognition on the ultrasonic echo signal interval based on the defect feature parameters and the optimal solution of the defect feature parameters, and determine the defect state of the steel plate based on the defect signal recognition result.

[0014] Third aspect, an embodiment of the present application further provides a computer device, which includes a memory and a processor, and a computer program is stored on the memory, and when the processor executes the computer program, the above method is implemented.

[0015] Fourth aspect, an embodiment of the present application further provides a computer-readable storage medium, and a computer program is stored on the storage medium, and when the computer program is executed by a processor, the above method can be implemented.

[0016] An embodiment of the present application provides an ultrasonic echo defect signal recognition method, device, equipment, and storage medium, which have the following advantages compared with the prior art:

[0017] By acquiring the ultrasonic echo signal of the steel plate, preprocessing the ultrasonic echo signal to obtain an ultrasonic echo signal interval; smoothing each target ultrasonic echo signal within the ultrasonic echo signal interval to obtain the corresponding smoothed ultrasonic echo signal, which can effectively suppress noise interference and reduce the influence of noise on the final recognition result; subsequently, based on the smoothed ultrasonic echo signal, determining all the maximum points and all the minimum points within the ultrasonic echo interval; based on all the maximum points and all the minimum points, extracting the defect feature parameters within the ultrasonic echo interval, which can comprehensively identify the important features in the echo signal, and is effective not only for defects with regular shapes but also for complex or irregularly shaped defects. Then, based on the Bayesian optimization technique, determining the optimal solution of the defect feature parameters corresponding to the defect feature parameters. The Bayesian optimization process avoids the fixed setting of specific parameters in traditional methods and has better adaptability and accuracy. Finally, based on the defect feature parameters and the optimal solution of the defect feature parameters, identifying the defect signal in the ultrasonic echo signal interval, and based on the defect signal recognition result, determining the defect state of the steel plate. The whole process not only reduces the interference of human factors on the detection result, but also enhances the noise suppression ability and improves the comprehensiveness and accuracy of defect recognition. BRIEF DESCRIPTION OF THE DRAWINGS

[0018] The accompanying drawings herein are incorporated into the specification and form a part of the specification, showing embodiments consistent with the present invention and, together with the specification, are used to explain the principles of the present invention.

[0019] To more clearly illustrate the technical solutions in the embodiments of the present invention or in the prior art, the following will briefly introduce the accompanying drawings required for the description of the embodiments or the prior art. Obviously, for those of ordinary skill in the art, other drawings can be obtained based on these drawings without creative efforts.

[0020] One or more embodiments are exemplarily illustrated by the pictures in the corresponding accompanying drawings. These exemplary illustrations do not limit the embodiments. Elements with the same reference numerals in the drawings represent similar elements. Unless otherwise stated, the drawings in the figures do not constitute a proportional limitation.

[0021] Figure 1 is a schematic flowchart of an embodiment of a method for identifying ultrasonic echo defect signals provided by the present application;

[0022] Figure 2 is a schematic structural diagram of an embodiment of a device for identifying ultrasonic echo defect signals provided by the present application;

[0023] Figure 3It is a schematic structural diagram of an electronic device provided by this application. Detailed implementation manners

[0024] To make the objectives, technical solutions and advantages of the embodiments of this application clearer, the technical solutions in the embodiments of this application will be clearly and completely described below with reference to the accompanying drawings in the embodiments of this application. Apparently, the described embodiments are some but not all of the embodiments of this application. All other embodiments obtained by those of ordinary skill in the art based on the embodiments of this application without creative efforts shall fall within the scope of protection of this application.

[0025] The following disclosure provides many different embodiments or examples for implementing different structures of this application. To simplify the disclosure of this application, components and settings of specific examples are described below. Of course, they are merely examples and are not intended to limit this application. In addition, this application may repeat reference numerals and / or letters in different examples. This repetition is for the purpose of simplification and clarity and does not in itself indicate the relationship between the various embodiments and / or settings discussed.

[0026] It should be understood that when used in this specification and the appended claims, the terms "include" and "comprise" indicate the presence of the described features, wholes, steps, operations, elements and / or components, but do not preclude the presence or addition of one or more other features, wholes, steps, operations, elements, components and / or their combinations.

[0027] It should also be understood that the terms used in this specification of this application are merely for the purpose of describing specific embodiments and are not intended to limit this application. As used in this specification of this application and the appended claims, unless the context clearly indicates otherwise, the singular forms "a", "an" and "the" are intended to include the plural forms.

[0028] It should be further understood that the term "and / or" used in this specification of this application and the appended claims refers to any combination and all possible combinations of one or more of the associated listed items, and includes these combinations.

[0029] As used in this specification and the appended claims, the term "if" can be interpreted as "when", "once", "in response to determining" or "in response to detecting" according to the context. Similarly, the phrase "if determined" or "if [the described condition or event] is detected" can be interpreted as meaning "once determined", "in response to determining", "once [the described condition or event] is detected" or "in response to detecting [the described condition or event]" according to the context.

[0030] Embodiment 1, seeFigure 1 , Figure 1 is a schematic flowchart of an embodiment of an ultrasonic echo defect signal recognition method provided by this application. As Figure 1 shown, this method includes steps 101 - 104, specifically as follows:

[0031] Step 101: Obtain the ultrasonic echo signal of the steel plate, and preprocess the ultrasonic echo signal to obtain an ultrasonic echo signal interval.

[0032] In one embodiment, the ultrasonic echo signal of the steel plate is obtained, where the ultrasonic echo signal includes an echo signal and a bottom wave signal.

[0033] Specifically, an ultrasonic wave signal is emitted to the steel plate through the probe of an ultrasonic flaw detection device, and the ultrasonic echo signal reflected by the steel plate is received. Among them, the ultrasonic echo signal includes an echo signal and a bottom wave signal, and the echo signal includes but is not limited to a primary echo signal and a secondary echo signal.

[0034] Specifically, the primary echo signal is the echo signal that is first reflected back to the probe when the ultrasonic wave signal encounters an interface of non-uniform media such as pores, cracks, and inclusions inside the steel plate after being emitted. The secondary echo signal is the echo signal that is reflected back to the probe after encountering the interface of non-uniform media again after a primary reflection inside the steel plate.

[0035] Specifically, the bottom wave signal is the signal formed when the ultrasonic wave signal propagates to the bottom surface of the steel plate and is reflected back; since the bottom surface of the steel plate is a relatively regular and stable interface, the bottom wave signal can reflect the basic propagation situation during the flaw detection process to a certain extent, and is often used as a reference signal for comparing and analyzing other echo signals to assist in judging the internal quality status of the steel plate.

[0036] In one embodiment, the amplitude difference between the amplitude of each echo signal and the amplitude of the bottom wave signal in the ultrasonic echo signal is calculated respectively. When it is determined that the amplitude differences corresponding to a continuous plurality of echo signals are greater than a preset amplitude difference threshold, an ultrasonic echo signal interval is determined based on the continuous plurality of echo signals, and the continuous plurality of echo signals are respectively used as the target ultrasonic echo signals within the ultrasonic echo signal interval.

[0037] Specifically, since the signal amplitude reflects the strength of the signal, by calculating the amplitude difference between the amplitude of each echo signal and the amplitude of the bottom wave signal, the intensity difference of each echo signal relative to the bottom wave signal can be quantified; and the intensity difference is related to whether there are defects inside the steel plate, because the existence of defects will change the propagation and reflection of ultrasonic waves, thereby affecting the signal amplitude of the echo signal.

[0038] Specifically, after calculating the amplitude difference corresponding to each echo signal in the ultrasonic echo signal, compare each amplitude difference with a preset amplitude difference threshold. If it is detected that the amplitude differences corresponding to multiple consecutive echo signals are all greater than the preset amplitude difference threshold, the signal interval where the multiple consecutive echo signals are located is used as the ultrasonic echo signal interval.

[0039] Step 102: Smooth each target ultrasonic echo signal in the ultrasonic echo signal interval to obtain the corresponding smoothed ultrasonic echo signal, and based on the smoothed ultrasonic echo signal, determine all the maximum points and all the minimum points in the ultrasonic echo interval.

[0040] In one embodiment, obtain the echo signal point sequence corresponding to each target ultrasonic echo signal in the ultrasonic echo signal interval, and assign a corresponding weight value to each echo signal point in the echo signal point sequence; based on the weight value, perform a weighted summation process on the echo signal point sequence to obtain a weighted summation value of the echo signal points, and perform a summation process on all the weight values to obtain a total weight; based on the weighted summation value of the echo signal points and the total weight, calculate the corresponding smoothed ultrasonic echo signal for each target ultrasonic echo signal in the ultrasonic echo signal interval.

[0041] Specifically, the echo signal point sequence corresponding to each target ultrasonic echo signal refers to the sequence that includes the target ultrasonic echo signal and all the historical echo signal points corresponding to the target ultrasonic echo signal, where all the historical echo signal points refer to multiple historical echo signal points before the target ultrasonic echo signal; the echo signal point sequence records the change situation of the target ultrasonic echo signal over a period of time.

[0042] Specifically, each echo signal point in the echo signal point sequence is sorted in ascending order according to its corresponding acquisition time.

[0043] Specifically, obtain a preset smoothing coefficient, and assign a corresponding weight value to each echo signal point in the echo signal point sequence based on the smoothing coefficient.

[0044] Since the magnitude of the weight value determines the influence degree of each echo signal point on the target ultrasonic echo signal in subsequent calculations; generally speaking, the echo signal points closer to the target ultrasonic echo signal have relatively larger weight values because they can better reflect the current characteristics of the target ultrasonic echo signal; while the echo signal points farther away have relatively smaller weight values. This way of weight assignment conforms to the actual situation because recent data often has a greater impact on the current state.

[0045] Specifically, substitute the smoothing coefficient into a preset weight value calculation formula to calculate the corresponding weight value assigned to each echo signal point in the echo signal point sequence. The weight value calculation formula is as follows: (1 - α) t-i , where α is the smoothing coefficient and t is the total number of echo signal points in the echo signal point sequence.

[0046] Specifically, based on the weight value corresponding to each echo signal point in the echo signal point sequence, perform a weighted summation process on the echo signal point sequence, that is, multiply the value of each echo signal point by its corresponding weight value, and then add all the products to obtain the weighted summation value of the echo signal points; doing so can comprehensively consider the numerical size of each point and its importance to the target signal, highlight the role of points with a greater impact on the target signal, and at the same time weaken the role of points with a smaller impact.

[0047] Specifically, when calculating the smoothed ultrasonic echo signal corresponding to each target ultrasonic echo signal using the weighted summation value of the echo signal points and the total weight, calculate the smoothed ultrasonic echo signal by dividing the weighted summation value of the echo signal points by the total weight; this calculation method uses the moving weighted exponential averaging technique to weight each data point with an exponentially decaying weight and normalizes the weights of all historical data, ensuring that as time goes by, the influence of earlier historical data gradually decreases, and at the same time avoiding the problem of weight imbalance, and realizing smoothing and denoising for each target ultrasonic echo signal within the ultrasonic echo signal interval.

[0048] In one embodiment, when smoothing each target ultrasonic echo signal within the ultrasonic echo signal interval, it is also possible to obtain the echo signal point sequence corresponding to each target ultrasonic echo signal within the ultrasonic echo signal interval, and substitute each echo signal point in the echo signal point sequence into a preset smoothing processing formula to obtain the corresponding smoothed ultrasonic echo signal; where the preset smoothing processing formula is as follows:

[0049]

[0050] In the formula, x i is the value of the i-th target ultrasonic echo signal, α is the smoothing coefficient that controls the attenuation rate, and y t is the value of the smoothed ultrasonic echo signal corresponding to the t-th target ultrasonic echo signal.

[0051] In one embodiment, traverse each smoothed ultrasonic echo signal within the ultrasonic echo signal interval, and during the traversal process, obtain the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal corresponding to the current smoothed ultrasonic echo signal.

[0052] Specifically, the first adjacent smoothed ultrasonic echo signal is the previous smoothed ultrasonic echo signal of the current smoothed ultrasonic echo signal, and the second adjacent smoothed ultrasonic echo signal is the next smoothed ultrasonic echo signal of the current smoothed ultrasonic echo signal.

[0053] In one embodiment, when the current smoothed ultrasonic echo signal is greater than the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal respectively, the current smoothed ultrasonic echo signal is taken as the maximum point.

[0054] In one embodiment, when the current smoothed ultrasonic echo signal is less than the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal respectively, the current smoothed ultrasonic echo signal is taken as the minimum point.

[0055] In one embodiment, until each smoothed ultrasonic echo signal in the ultrasonic echo signal interval is traversed, all the maximum points and all the minimum points in the ultrasonic echo interval are integrated.

[0056] In this embodiment, the moving weighted exponential averaging technique is used to smooth and denoise the echo signal, effectively reducing the interference of noise on signal analysis and ensuring the accurate extraction of signal features; this process not only considers the historical influence of data, but also avoids the problem of weight imbalance through the exponentially decaying weight distribution, thereby improving the stability and reliability of signal processing.

[0057] Step 103: Based on all the maximum points and all the minimum points, extract the defect feature parameters in the ultrasonic echo interval, and based on the Bayesian optimization technique, determine the optimal solution of the defect feature parameters corresponding to the defect feature parameters.

[0058] In one embodiment, the adjacent minimum points corresponding to each maximum point are obtained respectively, and for each maximum point and the adjacent minimum points corresponding to each maximum point, combination processing is performed to obtain a plurality of extreme point pairs.

[0059] Specifically, when obtaining the adjacent minimum points corresponding to each maximum point, the adjacent minimum points are the left adjacent minimum points.

[0060] Specifically, when the adjacent minimum points are the left adjacent minimum points, the extreme point pair corresponding to each maximum point is (left minimum point, maximum point).

[0061] Preferably, since all the maximum points and all the minimum points within the ultrasonic echo interval are arranged in the form of maximum point - minimum point - maximum point, or minimum point - maximum point - minimum point; therefore, first determine all the maximum points within the ultrasonic echo interval, and determine the previous minimum point corresponding to each maximum point. Based on all the maximum points, combine and pair them with their corresponding previous minimum points to obtain multiple pairs of extreme points.

[0062] In one embodiment, calculate the span values between the target maximum point and the target adjacent minimum point in each pair of extreme points respectively to obtain a two - dimensional span value sequence.

[0063] Specifically, calculate the distance between the target maximum point and the target adjacent minimum point in each pair of extreme points, and use the distance as the span value between the two.

[0064] Specifically, after obtaining the span values corresponding to all pairs of extreme points, sort all the span values based on the positions of the pairs of extreme points within the ultrasonic echo interval to obtain a two - dimensional span value sequence.

[0065] In one embodiment, obtain the target number of first pre - ultrasonic echo signals corresponding to the target maximum value in each pair of extreme points, and obtain the target number of first post - ultrasonic echo signals corresponding to the target minimum value in each pair of extreme points. Calculate the amplitude difference between the amplitude mean value of the first pre - ultrasonic echo signals corresponding to each pair of extreme points and the amplitude mean value of the first post - ultrasonic echo signals respectively to obtain a two - dimensional amplitude difference sequence.

[0066] Specifically, the target number g is obtained through Bayesian search.

[0067] Specifically, subtract the amplitude mean value of the first post - ultrasonic echo signals from the amplitude mean value of the first pre - ultrasonic echo signals respectively to obtain the amplitude difference corresponding to each pair of extreme points; sort all the amplitude differences based on the positions of the pairs of extreme points within the ultrasonic echo interval to obtain a two - dimensional amplitude difference sequence.

[0068] In one embodiment, obtain the first echo signal point sequence corresponding to each maximum point, calculate the first data change rate of adjacent ultrasonic echo signals in the first echo signal sequence respectively, determine the first data change rate range corresponding to each maximum point based on the first data change rate, and obtain a first two - dimensional data change rate range sequence based on the first data change rate range.

[0069] Specifically, the first echo signal point sequence corresponding to each maximum point includes the first four echo signal points and the last four echo signal points before and after the maximum value.

[0070] Specifically, calculate the first data change rate between adjacent two echo signal points in the first echo signal point sequence to obtain a plurality of first data change rates, and acquire the maximum value and the minimum value of the first data change rates among the plurality of first data change rates. Based on the maximum value and the minimum value of the first data change rates, determine the first data change rate range corresponding to each maximum point, and sort all the first data change rate ranges based on the position of the maximum point in the ultrasonic echo interval to obtain a first two-dimensional data change rate range sequence.

[0071] Specifically, when calculating the first data change rate between adjacent two echo signal points in the first echo signal point sequence, the first difference between the first echo signal point and the second echo signal point in the adjacent two echo signal points can be calculated, and the first difference is divided by the value of the first echo signal point to obtain the first data change rate between the adjacent two echo signal points.

[0072] In one embodiment, obtain the second echo signal point sequence corresponding to each maximum point, calculate the second data change rate between adjacent ultrasonic echo signals in the second echo signal sequence respectively, and based on the second data change rate, determine the second data change rate range corresponding to each maximum point, and obtain a second two-dimensional data change rate range sequence based on the second data change rate range.

[0073] Specifically, the second echo signal point sequence corresponding to each maximum point includes the first two echo signal points and the last two echo signal points before the maximum point.

[0074] Specifically, calculate the second data change rate between adjacent two echo signal points in the second echo signal point sequence to obtain a plurality of second data change rates, and acquire the maximum value and the minimum value of the second data change rates among the plurality of second data change rates. Based on the maximum value and the minimum value of the second data change rates, determine the second data change rate range corresponding to each maximum point, and sort all the second data change rate ranges based on the position of the maximum point in the ultrasonic echo interval to obtain a second two-dimensional data change rate range sequence.

[0075] Specifically, when calculating the second data change rate between adjacent two echo signal points in the second echo signal point sequence, the second difference between the first echo signal point and the second echo signal point in the adjacent two echo signal points can be calculated, and the second difference is divided by the value of the first echo signal point to obtain the second data change rate between the adjacent two echo signal points.

[0076] In one embodiment, the two-dimensional span value sequence h, the two-dimensional amplitude difference sequence f, the target quantity g, the first two-dimensional data change rate range sequence k1, and the second two-dimensional data change rate range sequence k2 are used as defect feature parameters within the ultrasonic echo interval.

[0077] In one embodiment, when determining the optimal solution of the defect feature parameters corresponding to the defect feature parameters based on the Bayesian optimization technique, a target function is constructed based on the defect feature parameters, and a Gaussian process regression model is constructed based on the defect feature parameters and the target function values corresponding to the defect feature parameters; based on a hybrid acquisition strategy with dynamically adjusted weights, an acquisition function is used to determine the evaluated defect feature parameters, and based on the evaluated defect feature parameters, model iteration processing is performed on the Gaussian process regression model until the target defect feature parameters corresponding to the minimum value of the target function are determined, and the target defect feature parameters are used as the optimal solution of the defect feature parameters corresponding to the defect feature parameters.

[0078] Specifically, the target function is f(a), where a = (h, g, f, k1, k2), and a is the defect feature parameter vector; the target function describes the misjudgment rate of the defect recognition model under a given feature combination.

[0079] Specifically, the Gaussian process regression model is as follows:

[0080] f(a) ∼ GP(m(a), k(a, a′));

[0081] where m(a) is the mean function of the target function, k(a, a′) is the kernel function, representing the similarity in the input space, f(a) is the target function value corresponding to the defect feature parameters, and GP(*) is the Gaussian process.

[0082] Specifically, in order to find the optimal combination of defect feature parameters, a hybrid acquisition strategy is adopted. During the optimization process, there is a trade-off between finding new, unexplored feature parameter spaces and continuing to search within existing, promising regions.

[0083] Specifically, the role of the acquisition function is to calculate an a for each possible defect feature parameter combination a based on the prediction and uncertainty of the Gaussian process regression model, and this a determines the potential value of the defect feature parameter combination a for minimizing the target function; among them, different acquisition functions have different ways of balancing exploration and exploitation, such as the expected improvement function, the upper confidence bound function, etc.

[0084] Specifically, by dynamically adjusting the weights of the acquisition function, it is possible to better balance exploration and exploitation at different stages of optimization; in the early stage of optimization, exploration may be more preferred to gain a broader understanding of the feature space; while in the later stage, when there is a certain understanding of the feature space, exploitation is more preferred to find the optimal solution using the known information.

[0085] Specifically, when using the acquisition function to determine the evaluation defect feature parameters, according to the current Gaussian process regression model and the acquisition function, calculate the next evaluation defect feature parameter a next , that is:

[0086] a next = argmin a Acquisition Function(f(a));

[0087] In the formula, Acquisition Function(*) is the acquisition function.

[0088] Specifically, evaluate the evaluation defect feature parameter a next to obtain the corresponding objective function value f(a next ), add the new evaluation defect feature parameter a next and the objective function value f(a next ) to the existing dataset, update the Gaussian process regression model, and the update process will adjust the mean function m(a) and the kernel function k(a,a′) according to the new data, making the prediction of the Gaussian process regression model for the objective function more accurate; continuously repeat the above steps, and update the Gaussian process regression model in each iteration, gradually approaching the global minimum of the objective function.

[0089] Specifically, as the number of iterations increases, the Gaussian process regression model is continuously updated, and the acquisition function will guide the search process towards a better direction. When the maximum number of iterations is reached or the change in the objective function value is less than a certain threshold, obtain the target defect feature parameter corresponding to the minimum value of the objective function, and use it as the optimal solution a′=(h′, g′, f′, k′1, k′2) of the defect feature parameter corresponding to the said defect feature parameter.

[0090] In this embodiment, by combining Bayesian optimization technology to optimize the defect recognition parameters, it is possible to automatically adjust and find the optimal solution during defect classification. By constructing a Gaussian process regression model, this method can effectively balance exploration and exploitation and quickly converge to the global optimal solution; this intelligent parameter optimization process improves the adaptability and recognition ability of the algorithm to different defect features, thereby improving the automation level and accuracy of steel plate flaw detection, reducing the dependence on manual intervention to set thresholds, and promoting the intelligent development of the industrial detection field.

[0091] Step 104: Based on the defect feature parameters and the optimal solution of the defect feature parameters, identify defect signals in the ultrasonic echo signal interval, and determine the defect state of the steel plate based on the defect signal identification result.

[0092] In one embodiment, the defect feature parameters include a two-dimensional span value sequence, a two-dimensional amplitude difference sequence, the number of targets, a first two-dimensional data change rate range sequence, and a second two-dimensional data change rate range sequence; the optimal solution of the defect feature parameters includes an optimal span value, an optimal number of targets, an optimal amplitude difference determined based on the optimal number of targets, an optimal first data change rate range, and an optimal second data change rate range.

[0093] In one embodiment, when any target span value in the two-dimensional span value sequence is greater than the optimal span value, any target amplitude difference in the two-dimensional amplitude difference sequence is greater than the optimal amplitude difference, any target first data change rate range in the first two-dimensional data change rate range sequence belongs to the optimal first data change rate range, and any target second data change rate range in the second two-dimensional data change rate range sequence belongs to the optimal second data change rate range, it is determined that there is a defect signal in the ultrasonic echo signal interval; wherein, the any target span value, the any target amplitude difference, the any target first data change rate range, and the any target second data change rate range are calculated from the same extreme point pair.

[0094] In one embodiment, when a defect signal is detected in the ultrasonic echo signal interval, the defect signal identification result is determined to be that there is a defect signal. At this time, the defect state of the steel plate is that there is a defect at the internal position of the steel plate where the ultrasonic echo signal interval is generated. Conversely, the defect state of the steel plate is that there is no defect at the internal position of the steel plate where the ultrasonic echo signal interval is generated.

[0095] Example 2, see Figure 2 , Figure 2 is a schematic structural diagram of an embodiment of an ultrasonic echo defect signal identification device provided by the present application. Corresponding to the above ultrasonic echo defect signal identification method, the present application also provides an ultrasonic echo defect signal identification device; the ultrasonic echo defect signal identification device includes a module for executing the above ultrasonic echo defect signal identification method, and the ultrasonic echo defect signal identification device can be configured in terminals such as desktop computers, tablet computers, laptop computers, etc. Specifically, the ultrasonic echo defect signal identification device includes a signal preprocessing module 201, a signal smoothing processing module 202, a Bayesian optimization module 203, and a defect signal identification module 204.

[0096] The signal preprocessing module 201 is configured to obtain the ultrasonic echo signal of the steel plate, preprocess the ultrasonic echo signal, and obtain an ultrasonic echo signal interval.

[0097] The signal smoothing processing module 202 is configured to perform smoothing processing on each target ultrasonic echo signal in the ultrasonic echo signal interval to obtain the corresponding smoothed ultrasonic echo signal, and determine all maximum points and all minimum points in the ultrasonic echo interval based on the smoothed ultrasonic echo signal.

[0098] The Bayesian optimization module 203 is configured to extract defect feature parameters in the ultrasonic echo interval based on all the maximum points and all the minimum points, and determine the optimal solution of the defect feature parameters corresponding to the defect feature parameters based on the Bayesian optimization technique.

[0099] The defect signal recognition module 204 is configured to perform defect signal recognition on the ultrasonic echo signal interval based on the defect feature parameters and the optimal solution of the defect feature parameters, and determine the defect state of the steel plate based on the defect signal recognition result.

[0100] In one embodiment, the signal preprocessing module 201 is configured to obtain the ultrasonic echo signal of the steel plate, preprocess the ultrasonic echo signal, and obtain an ultrasonic echo signal interval, specifically including: obtaining the ultrasonic echo signal of the steel plate, where the ultrasonic echo signal includes an echo signal and a bottom wave signal; respectively calculating the amplitude difference between the amplitude of each echo signal and the amplitude of the bottom wave signal in the ultrasonic echo signal, and when it is determined that the amplitude differences corresponding to a continuous plurality of echo signals are greater than a preset amplitude difference threshold, determining an ultrasonic echo signal interval based on the continuous plurality of echo signals, and respectively using the continuous plurality of echo signals as the target ultrasonic echo signals in the ultrasonic echo signal interval.

[0101] In one embodiment, the signal smoothing processing module 202 is configured to perform smoothing processing on each target ultrasonic echo signal in the ultrasonic echo signal interval to obtain the corresponding smoothed ultrasonic echo signal, specifically including: obtaining the echo signal point sequence corresponding to each target ultrasonic echo signal in the ultrasonic echo signal interval, and assigning a corresponding weight value to each echo signal point in the echo signal point sequence; performing weighted summation processing on the echo signal point sequence based on the weight value to obtain a weighted summation value of the echo signal points, and performing summation processing on all the weight values to obtain a total weight; calculating the corresponding smoothed ultrasonic echo signal for each target ultrasonic echo signal in the ultrasonic echo signal interval based on the weighted summation value of the echo signal points and the total weight.

[0102] In one embodiment, the signal smoothing processing module 202 is configured to determine all maximum points and all minimum points within the ultrasonic echo interval based on the smoothed ultrasonic echo signal. Specifically, it includes: traversing each smoothed ultrasonic echo signal within the ultrasonic echo signal interval, and during the traversal process, obtaining the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal corresponding to the current smoothed ultrasonic echo signal; when the current smoothed ultrasonic echo signal is greater than the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal respectively, taking the current smoothed ultrasonic echo signal as a maximum point; when the current smoothed ultrasonic echo signal is less than the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal respectively, taking the current smoothed ultrasonic echo signal as a minimum point; until the traversal is completed, integrating to obtain all maximum points and all minimum points within the ultrasonic echo interval.

[0103] In one embodiment, the Bayesian optimization module 203 is configured to extract defect feature parameters within the ultrasonic echo interval based on all the maximum points and all the minimum points. Specifically, it includes: respectively obtaining the adjacent minimum points corresponding to each maximum point, performing combination processing on each maximum point and the adjacent minimum points corresponding to each maximum point to obtain a plurality of extreme point pairs; respectively calculating the span values between the target maximum point and the target adjacent minimum point in each extreme point pair to obtain a two-dimensional span value sequence; obtaining the target number of first pre-ultrasonic echo signals corresponding to the target maximum value in each extreme point pair, and obtaining the target number of first post-ultrasonic echo signals corresponding to the target minimum value in each extreme point pair, respectively calculating the amplitude difference between the amplitude mean of the first pre-ultrasonic echo signals corresponding to each extreme point pair and the amplitude mean of the first post-ultrasonic echo signals to obtain a two-dimensional amplitude difference sequence; obtaining the first echo signal point sequence corresponding to each maximum point, respectively calculating the first data change rate of adjacent ultrasonic echo signals in the first echo signal sequence, and based on the first data change rate, determining the first data change rate range corresponding to each maximum point, and based on the first data change rate range, obtaining a first two-dimensional data change rate range sequence; obtaining the second echo signal point sequence corresponding to each maximum point, respectively calculating the second data change rate of adjacent ultrasonic echo signals in the second echo signal sequence, and based on the second data change rate, determining the second data change rate range corresponding to each maximum point, and based on the second data change rate range, obtaining a second two-dimensional data change rate range sequence; taking the two-dimensional span value sequence, the two-dimensional amplitude difference sequence, the target number, the first two-dimensional data change rate range sequence, and the second two-dimensional data change rate range sequence as the defect feature parameters within the ultrasonic echo interval.

[0104] In one embodiment, the Bayesian optimization module 203 is configured to determine an optimal solution of the defect feature parameters corresponding to the defect feature parameters based on Bayesian optimization technology, specifically including: constructing an objective function based on the defect feature parameters, and constructing a Gaussian process regression model based on the defect feature parameters and the objective function values corresponding to the defect feature parameters; determining evaluation defect feature parameters by using an acquisition function based on a hybrid acquisition strategy with dynamically adjusted weights, and performing model iteration processing on the Gaussian process regression model based on the evaluation defect feature parameters until an objective defect feature parameter corresponding to the minimum value of the objective function is determined, and taking the objective defect feature parameter as the optimal solution of the defect feature parameters corresponding to the defect feature parameters.

[0105] In one embodiment, the defect signal recognition module 204 is configured to perform defect signal recognition on the ultrasonic echo signal interval based on the defect feature parameters and the optimal solution of the defect feature parameters, specifically including: the defect feature parameters include a two-dimensional span value sequence, a two-dimensional amplitude difference sequence, a target number, a first two-dimensional data change rate range sequence, and a second two-dimensional data change rate range sequence; the optimal solution of the defect feature parameters includes an optimal span value, an optimal target number, an optimal amplitude difference determined based on the optimal target number, an optimal first data change rate range, and an optimal second data change rate range; when any target span value in the two-dimensional span value sequence is greater than the optimal span value, any target amplitude difference in the two-dimensional amplitude difference sequence is greater than the optimal amplitude difference, any target first data change rate range in the first two-dimensional data change rate range sequence belongs to the optimal first data change rate range, and any target second data change rate range in the second two-dimensional data change rate range sequence belongs to the optimal second data change rate range, it is determined that there is a defect signal in the ultrasonic echo signal interval; wherein, the any target span value, the any target amplitude difference, the any target first data change rate range, and the any target second data change rate range are calculated from the same pair of extreme points.

[0106] The above ultrasonic echo defect signal recognition device can implement the ultrasonic echo defect signal recognition method in the above method embodiment. The optional items in the above method embodiment are also applicable to this embodiment and will not be elaborated here.

[0107] As Figure 3 shown, Figure 3 FIG. is a schematic structural diagram of an electronic device provided by the present application; it includes a processor 111, a communication interface 112, a memory 113, and a communication bus 114, wherein the processor 111, the communication interface 112, and the memory 113 complete mutual communication through the communication bus 114, and the memory 113 is used to store a computer program.

[0108] In one embodiment of the present application, when the processor 111 executes the program stored on the memory 113, it implements the ultrasonic echo defect signal recognition method provided in any of the foregoing method embodiments.

[0109] Those of ordinary skill in the art can understand that all or part of the processes in the methods of the above embodiments can be completed by instructing relevant hardware through a computer program. This computer program can be stored in a storage medium, and this storage medium is a computer-readable storage medium. This computer program is executed by at least one processor in the computer system to implement the process steps of the above method embodiments.

[0110] Therefore, the embodiments of the present application also provide a computer-readable storage medium, on which a computer program is stored. When the computer program is executed by a processor, it implements the steps of the ultrasonic echo defect signal recognition method provided in any of the foregoing method embodiments.

[0111] The storage medium is a physical, non-transitory storage medium. For example, it can be various physical storage media such as a USB flash drive, a mobile hard disk, a read-only memory (ROM), a magnetic disk, or an optical disc that can store program codes. The computer-readable storage medium can be non-volatile or volatile.

[0112] Those of ordinary skill in the art can realize that the units and algorithm steps of the examples described in conjunction with the embodiments disclosed herein can be implemented by electronic hardware, computer software, or a combination of the two. To clearly illustrate the interchangeability of hardware and software, the components and steps of the examples have been generally described according to their functions in the above description. Whether these functions are executed in a hardware or software manner depends on the specific application and design constraints of the technical solution. Professional technicians can use different methods to implement the described functions for each specific application, but such implementation should not be considered to exceed the scope of the present application.

[0113] In several embodiments provided by the present application, it should be understood that the disclosed devices and methods can be implemented in other ways. For example, the device embodiments described above are merely illustrative. For example, the division of each unit is only a logical function division, and there can be other division methods in actual implementation. For example, multiple units or components can be combined or integrated into another system, or some features can be ignored or not executed.

[0114] The steps in the method of the embodiments of the present application can be adjusted in sequence, combined, and deleted according to actual needs. The units in the device of the embodiments of the present application can be combined, divided, and deleted according to actual needs. In addition, in each embodiment of the present application, each functional unit can be integrated in a processing unit, or each unit can exist physically alone, or two or more units can be integrated in one unit.

[0115] If the integrated unit is implemented in the form of a software functional unit and sold or used as an independent product, it can be stored in a storage medium. Based on this understanding, the technical solution of the present application, in essence, or the part that contributes to the prior art, or all or part of the technical solution can be embodied in the form of a software product. The computer software product is stored in a storage medium and includes several instructions for causing a computer device (which can be a personal computer, a terminal, or a network device, etc.) to execute all or part of the steps of the methods described in the embodiments of the present application.

[0116] In the above embodiments, the descriptions of the respective embodiments have their own emphases. For the parts not described in detail in a certain embodiment, reference can be made to the relevant descriptions of other embodiments.

[0117] Obviously, those skilled in the art can make various changes and modifications to the present application without departing from the spirit and scope of the present application. Thus, provided that these modifications and variations of the present application fall within the scope of the claims of the present application and their equivalent technologies, the present application also intends to include these changes and modifications.

[0118] As described above, the above are only the specific embodiments of the present application, but the protection scope of the present application is not limited thereto. Any person skilled in the art within the technical scope disclosed by the present application can easily think of various equivalent modifications or substitutions, and these modifications or substitutions should all be covered by the protection scope of the present application. Therefore, the protection scope of the present application shall be subject to the protection scope of the claims.

Claims

1. An ultrasonic echo defect signal recognition method, characterized in that Including: Obtain the ultrasonic echo signal of the steel plate, preprocess the ultrasonic echo signal to obtain an ultrasonic echo signal interval; Smooth each target ultrasonic echo signal in the ultrasonic echo signal interval to obtain the corresponding smoothed ultrasonic echo signal, and based on the smoothed ultrasonic echo signal, determine all the maximum points and all the minimum points in the ultrasonic echo interval; Based on all the maximum points and all the minimum points, extract the defect characteristic parameters in the ultrasonic echo interval, and based on the Bayesian optimization technique, determine the optimal solution of the defect characteristic parameters corresponding to the defect characteristic parameters; Based on the defect characteristic parameters and the optimal solution of the defect characteristic parameters, identify the defect signal in the ultrasonic echo signal interval, and based on the defect signal identification result, determine the defect state of the steel plate.

2. The method according to claim 1 above, characterized in that, Obtain the ultrasonic echo signal of the steel plate, preprocess the ultrasonic echo signal to obtain an ultrasonic echo signal interval, specifically including: Obtain the ultrasonic echo signal of the steel plate, wherein the ultrasonic echo signal includes an echo signal and a bottom echo signal; Calculate the amplitude difference between the amplitude of each echo signal and the amplitude of the bottom echo signal in the ultrasonic echo signal respectively. When it is determined that the amplitude differences corresponding to a continuous plurality of echo signals are greater than a preset amplitude difference threshold, based on the continuous plurality of echo signals, determine the ultrasonic echo signal interval, and respectively use the continuous plurality of echo signals as the target ultrasonic echo signals in the ultrasonic echo signal interval.

3. The method according to claim 1 above, characterized in that, Smooth each target ultrasonic echo signal in the ultrasonic echo signal interval to obtain the corresponding smoothed ultrasonic echo signal, specifically including: Obtain the echo signal point sequence corresponding to each target ultrasonic echo signal in the ultrasonic echo signal interval, and assign a corresponding weight value to each echo signal point in the echo signal point sequence; Based on the weight value, perform a weighted summation process on the echo signal point sequence to obtain a weighted summation value of the echo signal points, and perform a summation process on all the weight values to obtain a total weight; Based on the weighted summation value of the echo signal points and the total weight, calculate the corresponding smoothed ultrasonic echo signal for each target ultrasonic echo signal in the ultrasonic echo signal interval.

4. The method according to claim 1 above, characterized in that, Based on the smoothed ultrasonic echo signal, determine all the maximum points and all the minimum points in the ultrasonic echo interval, specifically including: Traverse each smoothed ultrasonic echo signal in the ultrasonic echo signal interval, and during the traversal process, obtain the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal corresponding to the current smoothed ultrasonic echo signal; When the current smoothed ultrasonic echo signal is respectively greater than the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal, use the current smoothed ultrasonic echo signal as the maximum point; When the current smoothed ultrasonic echo signal is respectively smaller than the first adjacent smoothed ultrasonic echo signal and the second adjacent smoothed ultrasonic echo signal, the current smoothed ultrasonic echo signal is used as the minimum point; Until the traversal is completed, all the maximum points and all the minimum points within the ultrasonic echo interval are integrated.

5. The method according to claim 1 above, characterized in that, Based on all the maximum points and all the minimum points, defect characteristic parameters within the ultrasonic echo interval are extracted, specifically including: Respectively obtain the adjacent minimum points corresponding to each maximum point, and perform combination processing on each maximum point and the adjacent minimum points corresponding to each maximum point to obtain a plurality of extreme point pairs; Respectively calculate the span values between the target maximum point and the target adjacent minimum point in each extreme point pair to obtain a two-dimensional span value sequence; Obtain the target number of first pre-ultrasonic echo signals corresponding to the target maximum value in each extreme point pair, and obtain the target number of first post-ultrasonic echo signals corresponding to the target minimum value in each extreme point pair. Respectively calculate the amplitude difference between the amplitude mean value of the first pre-ultrasonic echo signals corresponding to each extreme point pair and the amplitude mean value of the first post-ultrasonic echo signals to obtain a two-dimensional amplitude difference sequence; Obtain the first echo signal point sequence corresponding to each maximum point, and respectively calculate the first data change rate of adjacent ultrasonic echo signals in the first echo signal sequence. Based on the first data change rate, determine the first data change rate range corresponding to each maximum point, and based on the first data change rate range, obtain a first two-dimensional data change rate range sequence; Obtain the second echo signal point sequence corresponding to each maximum point, and respectively calculate the second data change rate of adjacent ultrasonic echo signals in the second echo signal sequence. Based on the second data change rate, determine the second data change rate range corresponding to each maximum point, and based on the second data change rate range, obtain a second two-dimensional data change rate range sequence; The two-dimensional span value sequence, the two-dimensional amplitude difference sequence, the target number, the first two-dimensional data change rate range sequence, and the second two-dimensional data change rate range sequence are used as the defect characteristic parameters within the ultrasonic echo interval.

6. The method according to claim 1 above, characterized in that, Based on the Bayesian optimization technique, determine the optimal solution of the defect characteristic parameters corresponding to the defect characteristic parameters, specifically including: Based on the defect characteristic parameters, construct an objective function, and based on the defect characteristic parameters and the objective function values corresponding to the defect characteristic parameters, construct a Gaussian process regression model; Based on a hybrid acquisition strategy with dynamically adjusted weights, use an acquisition function to determine the evaluation of the defect characteristic parameters, and based on the evaluation of the defect characteristic parameters, perform model iteration processing on the Gaussian process regression model until the target defect characteristic parameters corresponding to the minimum value of the objective function are determined, and use the target defect characteristic parameters as the optimal solution of the defect characteristic parameters corresponding to the defect characteristic parameters.

7. The method according to claim 5 above, characterized in that, Based on the defect characteristic parameters and the optimal solution of the defect characteristic parameters, perform defect signal recognition on the ultrasonic echo signal interval, specifically including: The defect characteristic parameters include a two-dimensional span value sequence, a two-dimensional amplitude difference sequence, the number of targets, a first two-dimensional data change rate range sequence, and a second two-dimensional data change rate range sequence; The optimal solution of the defect characteristic parameters includes an optimal span value, an optimal number of targets, an optimal amplitude difference determined based on the optimal number of targets, an optimal first data change rate range, and an optimal second data change rate range; When any target span value in the two-dimensional span value sequence is greater than the optimal span value, any target amplitude difference in the two-dimensional amplitude difference sequence is greater than the optimal amplitude difference, any target first data change rate range in the first two-dimensional data change rate range sequence belongs to the optimal first data change rate range, and any target second data change rate range in the second two-dimensional data change rate range sequence belongs to the optimal second data change rate range, it is determined that there is a defect signal in the ultrasonic echo signal interval; Wherein, the any target span value, the any target amplitude difference, the any target first data change rate range, and the any target second data change rate range are calculated from the same extreme point pair.

8. An ultrasonic echo defect signal recognition device, characterized in that, Including: A signal preprocessing module, a signal smoothing processing module, a Bayesian optimization module, and a defect signal identification module; Wherein, the signal preprocessing module is configured to obtain the ultrasonic echo signal of the steel plate, preprocess the ultrasonic echo signal, and obtain an ultrasonic echo signal interval; The signal smoothing processing module is configured to perform smoothing processing on each target ultrasonic echo signal in the ultrasonic echo signal interval to obtain the corresponding smoothed ultrasonic echo signal, and based on the smoothed ultrasonic echo signal, determine all the maximum points and all the minimum points in the ultrasonic echo interval; The Bayesian optimization module is configured to extract the defect characteristic parameters in the ultrasonic echo interval based on all the maximum points and all the minimum points, and determine the optimal solution of the defect characteristic parameters corresponding to the defect characteristic parameters based on the Bayesian optimization technique; The defect signal identification module is configured to identify the defect signal in the ultrasonic echo signal interval based on the defect characteristic parameters and the optimal solution of the defect characteristic parameters, and determine the defect state of the steel plate based on the defect signal identification result.

9. A computer device, characterized in that, The computer device includes a memory and a processor, and a computer program is stored on the memory. When the processor executes the computer program, the method described in any one of claims 1-7 is implemented.

10. A computer-readable storage medium, characterized in that, The storage medium stores a computer program, and when the computer program is executed by the processor, the method described in any one of claims 1-7 can be implemented.

Images