System for detecting small defects in the interior of mechanical materials based on ultrasound

Abstract

The application discloses a system for detecting tiny defects in mechanical materials based on ultrasonic waves, and relates to the field of ultrasonic detection.The system comprises a data acquisition module, an envelope line generation module, a feature extraction module, a first judgment module and a second judgment module.The system generates envelope lines representing energy distribution by acquiring ultrasonic echo sequences, and extracts features such as peak amplitude, position and echo width.The system judges whether the peak amplitude is lower than a dynamic noise threshold.If the peak amplitude is lower than the dynamic noise threshold, it is determined that there is no defect.Otherwise, the echo width ratio is calculated.If the ratio is in a preset interval, a typical defect is output.If the ratio is lower than the lower limit of the interval, the gradient change of the peak position is further analyzed to identify tiny defects.If the ratio is higher than the upper limit, local maximum points and the minimum distance between the local maximum points and the peak position are found to distinguish multiple defects or a single large defect.The system realizes multi-dimensional accurate identification and classification of defects in materials.

Description

TECHNICAL FIELD

[0001] The present application relates to the field of ultrasonic detection, in particular to a system for detecting micro-defects in mechanical materials based on ultrasonic detection. BACKGROUND

[0002] As the core means of industrial non-destructive testing, ultrasonic detection technology plays an irreplaceable role in the identification and evaluation of internal defects in mechanical materials. Its basic principle is to emit ultrasonic waves to the material and receive the reflected echoes, and to infer the existence, position and shape of the defects according to the characteristic changes of the echo signals.

[0003] With the continuous improvement of the requirements for material integrity in high-end manufacturing industry, especially the increasing demand for detection of micron-level micro-defects, the existing detection systems face severe challenges in sensitivity and accuracy. When the defects present complex shapes, the diversity of defect shapes leads to distortion of echo width characteristics, and the existing solutions do not introduce multi-dimensional indicators for comprehensive verification, which significantly reduces the reliability of the detection results under noise fluctuations or material non-uniformity conditions.

[0004] Therefore, how to construct a detection system that can accurately extract weak signal features from strong noise background and realize intelligent discrimination of defect types has become a key bottleneck that needs to be broken through in current ultrasonic detection technology. SUMMARY

[0005] In view of the deficiencies of the prior art, the present application provides a system for detecting micro-defects in mechanical materials based on ultrasonic detection.

[0006] In order to achieve the above-mentioned purpose, the technical scheme of the present application is as follows:

[0007] In a first aspect, the present application discloses a system for detecting micro-defects in mechanical materials based on ultrasonic detection, comprising:

[0008] A data acquisition module for acquiring an ultrasonic echo sequence, the ultrasonic echo sequence including amplitude values of a plurality of sampling points;

[0009] An envelope line generation module for generating an envelope line from the ultrasonic echo sequence through moving average processing, the envelope line representing the energy distribution of the echoes;

[0010] A feature extraction module for extracting peak amplitude, peak position and echo width from the envelope line;

[0011] A first judgment module for judging whether the peak amplitude is less than a noise threshold, if yes, outputting a no-defect result, otherwise calculating the echo width ratio according to the ratio of the echo width and the standard echo width;

[0012] A second judging module is configured to judge whether the echo width ratio is in a preset range interval, and if yes, output a typical defect result;

[0013] Otherwise, when the echo width ratio is less than the lower limit of the range interval, a gradient change of the peak position is calculated according to the envelope line; it is judged whether the gradient change is greater than a preset gradient threshold, and if yes, a micro defect result is output, otherwise, a no defect result is output;

[0014] When the echo width ratio is greater than the upper limit of the range interval, all local maximum points are searched according to the envelope line, and a minimum distance between the local maximum points and the peak position is calculated; it is judged whether the minimum distance is less than a preset distance threshold, and if yes, a multiple defect result is output, otherwise, a single large defect result is output.

[0015] In a second aspect, the present application discloses a method for detecting micro defects in mechanical materials based on ultrasonic waves, comprising the following steps:

[0016] An ultrasonic echo sequence is obtained, and the ultrasonic echo sequence comprises amplitude values of a plurality of sampling points;

[0017] An envelope line is obtained by moving average processing according to the ultrasonic echo sequence, and the envelope line represents the energy distribution of the echo;

[0018] A peak amplitude, a peak position and an echo width are extracted according to the envelope line;

[0019] It is judged whether the peak amplitude is less than a noise threshold, and if yes, a no defect result is output, otherwise, an echo width ratio is calculated according to the ratio of the echo width and a standard echo width;

[0020] It is judged whether the echo width ratio is in a preset range interval, and if yes, a typical defect result is output;

[0021] Otherwise, when the echo width ratio is less than the lower limit of the range interval, a gradient change of the peak position is calculated according to the envelope line; it is judged whether the gradient change is greater than a preset gradient threshold, and if yes, a micro defect result is output, otherwise, a no defect result is output;

[0022] When the echo width ratio is greater than the upper limit of the range interval, all local maximum points are searched according to the envelope line, and a minimum distance between the local maximum points and the peak position is calculated; it is judged whether the minimum distance is less than a preset distance threshold, and if yes, a multiple defect result is output, otherwise, a single large defect result is output.

[0023] Compared with the prior art, the present application has the following beneficial effects:

[0024] 1. Utilizing echo width ratio, local maximum point distribution, energy concentration, etc. Multi-dimensional characteristics for comprehensive judgment, the system can accurately distinguish different types such as typical defects, small defects, multiple defects aggregation and single large defect, avoid misclassification caused by complex defect morphology, and improve the accuracy of detection results;

[0025] 2. Through the envelope generation of adaptive window size, effectively suppress high-frequency noise interference, strengthen the representation of echo energy distribution, provide high-fidelity data basis for peak amplitude, position and echo width extraction, thereby improve the reliability of defect characteristics;

[0026] 3. Through the iteration processing unit and defect confirmation mechanism, the system can dynamically adjust parameters to cope with different detection environments, ensure that it can still output reliable results under low signal-to-noise ratio or complex working conditions, and enhance the applicability and durability of the overall detection system;

[0027] 4. Through the dynamic noise threshold mechanism and gradient change analysis, the system can effectively distinguish small defect signals with low amplitude from environmental noise, significantly reduce the missed detection rate, and maintain stable performance especially under noise fluctuation or material non-uniformity conditions. BRIEF DESCRIPTION OF DRAWINGS

[0028] In order to more clearly illustrate the technical solutions of the embodiments of the present application or the prior art, the following will briefly introduce the drawings needed to be used in the embodiment or prior art description. Obviously, the drawings in the following description are only some embodiments of the present application, and for those skilled in the art, other drawings can also be obtained without creative labor on the basis of these drawings.

[0029] Figure 1 It is the overall block diagram of the system of the first embodiment of the present application;

[0030] Figure 2 It is the overall block diagram of the method of the second embodiment of the present application;

[0031] Figure 3 It is the execution flow chart of the method of the second embodiment of the present application. DETAILED DESCRIPTION

[0032] The technical solutions of the present application will be described in detail below in conjunction with the embodiments. Obviously, the described embodiments are only some of the embodiments of the present application, not all. Based on the embodiments in the present application, all other embodiments obtained by those skilled in the art without creative labor are within the scope of protection of the present application.

[0033] Summary of the application: In the traditional existing ultrasonic detection technology, the echo signal of the micro defect inside the mechanical material is easily disturbed by the environmental noise due to the low amplitude value, which leads to the fact that the amplitude comparison method based on fixed threshold value cannot effectively distinguish the defect signal from the noise; at the same time, when the defect distribution is uneven or the shape is abnormal, the echo waveform presents nonlinear change characteristics, the existing technology lacks multi-dimensional correlation analysis of echo width, gradient change and local maximum point distribution, and it is difficult to realize accurate classification of typical defects, micro defects, multiple defect aggregation and single large defect, thereby causing the reliability of the detection result to decrease and the classification accuracy to decrease.

[0034] For example, in the ultrasonic detection process of titanium alloy material of aero-engine turbine blade, the echo signal amplitude of micro crack is lower than the noise threshold value, which is judged as no defect signal; when multiple micro cracks gather in a local area, the ratio of echo width to standard echo width exceeds the upper limit of the preset range interval, but the existing system only compares the width with the preset standard value directly, and mistakenly identifies multiple defects as a single large defect, which leads to the fact that the defect evaluation result does not match the actual material state, and affects the accuracy of subsequent maintenance decision.

[0035] If the above problems are not solved, the missed detection of micro defects will increase the risk of sudden failure of materials in service process, and the misclassification of defect type may cause unnecessary replacement of parts or ignore potential safety hazards, which seriously damages the credibility of the detection system and the safety of industrial application.

[0036] Embodiment one:

[0037] As shown in Figure 1 The system for detecting micro defects inside mechanical materials based on ultrasonic waves comprises:

[0038] A data acquisition module is configured to acquire an ultrasonic echo sequence, wherein the ultrasonic echo sequence comprises amplitude values of a plurality of sampling points.

[0039] Specifically, first, high-frequency ultrasonic pulses are emitted to the mechanical material to be detected according to the deployed ultrasonic probe, which is usually composed of a piezoelectric transducer with a preset working frequency according to the material type and detection requirements, for example, a frequency in the range of 1 MHz to 10 MHz is often used in metal components. When the transmitted ultrasonic waves propagate in the material, they will produce reflected echoes when encountering defects or boundaries. These echoes are captured by the same probe or a receiving probe, forming an analog electrical signal. Then, the analog echo signal goes through a signal conditioning stage, including amplification and filtering processing. Amplification is achieved through a programmable gain amplifier to adjust the signal amplitude to the dynamic range suitable for analog-to-digital conversion; filtering uses a band-pass filter to suppress high-frequency noise and low-frequency interference, and to retain the effective echo frequency band. The conditioned analog signal is input into an analog-to-digital converter for digitization according to the preset sampling frequency. The sampling frequency needs to meet the Nyquist theorem, i.e., at least twice the center frequency of the ultrasonic wave, to ensure signal integrity. The analog-to-digital converter converts the continuous analog signal into a discrete digital sequence, outputting an echo sequence wherein represents the amplitude value of the th sampling point, is the sequence length, which is determined by the sampling frequency and the detection time window.

[0040] For example, when detecting an aluminum alloy component, a probe with a center frequency of 5 MHz is used, the sampling frequency is set to 25 MHz, and the detection time window is 100 microseconds, then the echo sequence length sampling points. The amplitude value of each sampling point is normalized to [-1, 1] for unified processing.

[0041] An envelope generation module is configured to generate an envelope line from the ultrasonic echo sequence through moving average processing, the envelope line representing the energy distribution of the echo.

[0042] The envelope generation module performs moving average processing on the echo sequence to obtain an envelope line representing the energy distribution of the echo, which effectively suppresses random noise interference and makes the energy distribution characteristics clearer.

[0043] Specifically, the purpose of extracting the envelope line is to remove high-frequency noise from the original echo sequence and highlight the energy distribution characteristics of the echo for subsequent feature extraction and defect analysis. This process is based on the moving average method, which calculates the average value of the absolute amplitude of the region near each sampling point in the echo sequence through a sliding window, thereby generating a smooth envelope line sequence. Specifically, first, obtain the digital echo sequence wherein represents the amplitude value of the th sampling point, which is used as input data. Then, define a fixed window size (typically an odd number to ensure symmetry), and calculate the half-window value For each sampling point in the sequence , extract the subsequence from , calculate the average of the absolute values of the amplitudes of all sampling points in the subsequence, and obtain the corresponding point of the envelope . This process is carried out point by point until the complete envelope sequence is generated Through this moving average processing, the rapid fluctuations in the original echo sequence are smoothed, while the overall energy trend of the echo is preserved, laying a foundation for subsequent peak amplitude, echo width, and other feature extraction.

[0044] The mathematical expression for envelope calculation is as follows:

[0045]

[0046] where, represents the value of the th point in the envelope, represents the fixed window size (positive integer), represents the half-window value (i.e. ), represents the amplitude value of the th sampling point in the echo sequence, represents the total length of the echo sequence.

[0047] Exemplarily, assume that the echo sequence contains 10 sampling points with values , and the window size . For the first point , the window covers indices 0 to 2 (assuming the index starts from 1, in practice, the boundary needs to be handled, for example, filled with zeros), but usually the index starts from 1, so when calculating, it needs to be adjusted: for , the window is , but is invalid, so in actual application, the boundary point may be handled by padding or reducing the window. For simplicity, this example ignores the boundary effect and only calculates the valid points. For example, for , the window is , the absolute value of the amplitude is , and the average value is . Similarly, for , the window is , the value is , and the average value is . Through point-by-point calculation, the final envelope sequence is obtained. The window size can be adjusted according to the actual sampling rate and noise level to balance the smoothness and detail preservation.

[0048] Further, the application proposes that the fixed window size in the envelope calculation process is determined according to the sampling frequency and the center frequency of the ultrasonic wave.

[0049] The fixed window refers to a set of sampling points used to limit the calculation range of local energy in the moving average processing, which can be implemented by using a symmetric rectangular window or a weighted window function, and the purpose is to isolate high-frequency noise interference and focus on the effective energy area of the signal. The determination of the fixed window size refers to dynamically calculating the window size based on the sampling frequency and the center frequency of the ultrasonic wave signal, which can be implemented by using a frequency domain matching algorithm or a preset mapping relationship, and the purpose is to adapt the window width to the signal main frequency bandwidth, avoiding defect feature blurring caused by too wide or introducing environmental noise caused by too narrow.

[0050] Specifically, the window size is dynamically determined according to the ratio of the sampling frequency to the center frequency, so that the window width is accurately matched with the effective frequency band of the ultrasonic wave signal, which prevents the smoothing and weakening of small defect features caused by too large window, and avoids the error inclusion of high-frequency noise in energy calculation caused by too small window. The specific calculation formula is: ; wherein, is the initial fixed window size, is the ratio of the sampling frequency to the center frequency, is a preset proportion coefficient.

[0051] Through the above scheme, the application significantly reduces the interference of environmental noise on envelope generation, improves the extraction accuracy of small defect weak signals, effectively avoids feature distortion and missed detection phenomenon, and enhances the reliability of defect detection.

[0052] The feature extraction module is configured to extract a peak amplitude, a peak position, and an echo width according to the envelope.

[0053] Further, the extraction process of the peak amplitude and the peak position includes: traversing all sampling points in the envelope, finding the maximum amplitude value as the peak amplitude, and recording the index of the maximum amplitude value in the ultrasonic echo sequence as the peak position; the extraction process of the echo width includes: taking a preset proportion of the peak amplitude as a width threshold, counting the number of sampling points in the envelope whose amplitude values continuously exceed the width threshold, and taking the number of sampling points as the echo width.

[0054] In practical applications, the traversal of all sampling points in the envelope line refers to global scanning of the envelope line data, which can be implemented in a sequential traversal or parallel computing manner to avoid misjudgment caused by local search; wherein the maximum amplitude value as the peak amplitude refers to selecting the global maximum value in the envelope line, which can be determined based on a numerical comparison algorithm, and accurately reflects the highest energy point of the echo signal; specifically, the peak position refers to the time sequence position of the maximum amplitude value in the ultrasonic echo sequence, which can be recorded as a sampling point index, providing an accurate time reference for subsequent gradient change calculation; taking the preset proportion of the peak amplitude as the width threshold refers to multiplying the peak amplitude by a fixed proportion coefficient to obtain the threshold, which can use a proportion coefficient between 50% and 70%, which can dynamically associate the threshold with the current signal strength, solving the failure problem of the fixed threshold under weak signal.

[0055] Specifically, the scheme of the present application determines the peak amplitude and position by globally traversing the envelope line, ensuring sensitive capture of weak defect signals in a noisy environment and avoiding peak misjudgment caused by local interference; at the same time, the dynamic width threshold based on the proportion of the peak amplitude makes the measurement process adaptive to different signal strengths, and the threshold can be scaled accordingly when the defect size changes; combined with the width statistical method with continuity constraint, random noise interference is effectively excluded, and the echo width feature truly reflects the defect morphology, thereby cooperatively realizing accurate quantization and reliable identification of small defect signals.

[0056] Through the above scheme, the present application effectively solves the problem that the peak recognition of small defects under noise interference is easily affected by random fluctuations, and the measurement of echo width can adapt to the dynamic characteristics of different defect signals, avoiding the missed detection of small defect signals due to noise, while improving the accuracy of echo width measurement under complex defect morphology, thereby accurately distinguishing small defects from typical defects or noise interference.

[0057] The first judgment module is used for judging whether the peak amplitude is smaller than the noise threshold, and if so, outputting a no defect result, otherwise, calculating the echo width ratio according to the ratio of the echo width and the standard echo width;

[0058] Specifically, first, the envelope line sequence is obtained and the peak amplitude is determined therefrom, which represents the maximum strength of the echo signal. Then, the system compares the peak amplitude with the noise threshold, wherein the noise threshold is dynamically adjusted based on the statistical characteristics of historical no defect samples or a preset fixed value. If the peak amplitude is smaller than the noise threshold, it is determined as no defect, and a no defect result is directly output. This step effectively excludes the interference of low amplitude noise through amplitude preliminary screening, avoiding unnecessary subsequent processing.

[0059] If the peak amplitude is not less than the noise threshold, the system enters the defect type analysis stage and calculates the echo width ratio. The echo width ratio is calculated by comparing the actual echo width with the standard echo width. The actual echo width is extracted from the envelope and defined as the number of consecutive sampling points in the envelope whose amplitude exceeds a preset proportional threshold based on the peak amplitude. The standard echo width is a reference value preset according to the material type and ultrasonic probe frequency, used to normalize the influence of different detection conditions. The echo width ratio quantifies the relative deviation of the echo duration, thereby helping to distinguish defect types: for example, when the echo width ratio is close to 1, it indicates that the echo duration is consistent with the standard condition, possibly corresponding to a typical defect; while a value deviating from 1 may indicate a small or large defect. Calculating the echo width ratio is necessary because relying solely on peak amplitude may not be sufficient to identify defect morphology, as different defects (such as cracks or pores) exhibit significant differences in echo width. Introducing the width ratio enhances the accuracy of defect classification.

[0060] The formula for calculating the echo width ratio is as follows:

[0061]

[0062] in, Indicates the echo width ratio. This represents the actual echo width, i.e., the amplitude in the envelope exceeding the threshold. Number of consecutive sampling points ( It is usually set to a preset percentage of the peak amplitude, for example ),and This indicates the standard echo width, pre-calibrated based on material properties and detection parameters.

[0063] The second judgment module is used to determine whether the echo width ratio is within a preset range; if so, it outputs typical defect results.

[0064] Otherwise, when the echo width ratio is less than the lower limit of the range, the gradient change of the peak position is calculated based on the envelope; it is determined whether the gradient change is greater than the preset gradient threshold. If it is, the result of minor defects is output; otherwise, the result of no defects is output.

[0065] When the echo width ratio is greater than the upper limit of the range, find all local maxima points according to the envelope, calculate the minimum distance between the local maxima point and the peak position, and determine whether the minimum distance is less than the preset distance threshold. If it is, output multiple defect results; otherwise, output a single large defect result.

[0066] Further, the process of calculating the gradient change of the peak position according to the envelope line comprises: calculating the absolute value of the difference of the envelope line at two adjacent sampling points of the peak position, and taking the absolute value of the difference as the gradient change; the process of finding all local maximum points according to the envelope line comprises: traversing each sampling point in the envelope line, judging whether the value corresponding to the sampling point in the envelope line is greater than the values corresponding to the previous sampling point and the next sampling point in the envelope line, and if so, taking the sampling point as a local maximum point.

[0067] The gradient change can be understood as a local change rate measure of the signal at the peak position, which can be realized by the absolute value of the difference of adjacent sampling points, and can be quantified by using generalized forms such as forward difference or backward difference, highlighting the rapid energy decay characteristic of the echo of the micro defect, and avoiding the blurring of details caused by a too wide window; the local maximum point can be understood as a local peak feature in the energy distribution, which can be identified by using parallel implementation modes such as sliding window comparison or dynamic threshold judgment, and can be specifically judged based on the relative relationship between the sampling point and its neighborhood, to stably extract the defect position information from the smoothed energy distribution and reduce the random fluctuation interference caused by noise.

[0068] Specifically, the scheme of the present application provides a stable basis for gradient change calculation and local maximum point finding through the smoothing property of the envelope line, so that the local feature change of the signal can be accurately captured in a noisy environment. The envelope line after moving average processing suppresses the high-frequency noise in the original echo, and the difference operation of adjacent sampling points can sensitively reflect the true steep change at the peak value; at the same time, the comparison mechanism based on the local neighborhood ensures that the maximum value point detection only depends on the morphological features of the signal itself, avoiding the subjectivity of global threshold setting. This design enables the system to reliably distinguish micro defects from noise fluctuations under low signal-to-noise ratio conditions, and accurately identify the real defect position under complex defect morphology, thereby forming an organic cooperation with the envelope line generation module and the feature extraction module in the preposed claims, and strengthening the robustness of defect classification.

[0069] Through the above technical scheme, the present application effectively reduces the interference of noise on the calculation of gradient change, avoids the generation of false local maximum points, and thereby improves the detection rate of micro defects and the classification accuracy of complex defect types in a low signal-to-noise ratio environment.

[0070] The system can reliably identify micro defect signals in a noisy environment through the fine analysis of the envelope line energy distribution and the synergistic effect of multi-level judgment logic, avoid the problem of missed detection caused by noise flooding, and accurately distinguish different types such as typical defects, micro defects, multiple defect aggregation and single large defect through the standardized processing of echo width ratio and the comprehensive judgment of gradient change and local maximum point, thereby significantly improving the accuracy of defect identification and the reliability of classification.

[0071] In some embodiments of the present application, in the implementation process, due to the lack of further verification mechanism for defect signals, when the echo width ratio is abnormal, the system may misjudge noise as a small defect or incorrectly identify multiple defects, resulting in reduced reliability of the detection results. In particular, the determination of small defects is easily affected by false gradient changes under noise interference, and the determination of multiple defects is easily affected by false local maximum points caused by noise.

[0072] To this end, the second judgment module further includes a defect confirmation unit for confirming the determined defect result, including:

[0073] The total energy of the envelope and the peak area energy are calculated, and the peak area energy is calculated in a fixed window centered on the peak position;

[0074] The energy concentration is calculated by ratio calculation according to the peak area energy and the total energy;

[0075] When the echo width ratio is less than the lower limit of the range interval, and the gradient change is greater than the gradient threshold, it is further judged whether the energy concentration is greater than a preset energy concentration threshold. If yes, the small defect result is confirmed, otherwise the defect result is confirmed by performing iterative processing;

[0076] When the echo width ratio is greater than the upper limit of the range interval, and multiple defect results are determined, only the local maximum point with an energy concentration greater than the energy concentration threshold is output as a real defect.

[0077] The defect confirmation unit refers to an additional processing module for verifying the defect determination result, which can be implemented by software algorithm or special hardware circuit, and its purpose is to strengthen the reliability evaluation of the defect signal through multi-dimensional energy analysis. The total energy of the envelope refers to the cumulative sum of the energy values of all sampling points in the envelope, which can be calculated by summing the square of the amplitudes of each sampling point in the envelope, and quantifies the overall energy level of the entire echo signal. The peak area energy refers to the total energy of the envelope in the local area centered on the peak position, which can be determined by a fixed window or an adaptive window to focus on the energy distribution characteristics of the defect core area. The energy concentration refers to the ratio of the peak area energy to the total energy of the envelope, which can be understood as an index of the concentration of energy in the peak area. Its purpose is to distinguish real defects from noise interference, because real defects usually exhibit higher energy concentration in local areas. Iterative processing refers to the process of re-executing part of the detection process by adjusting parameters, which can be implemented by dynamically adjusting the window size or optimizing the threshold, providing a dynamic verification path when the initial judgment is uncertain.

[0078] Specifically, the total energy of the envelope and the energy of the peak region centered on the peak position are calculated first to obtain the energy concentration index. When the echo width ratio is less than the lower limit of the range interval and the gradient change is greater than the gradient threshold, the energy concentration threshold is used for secondary verification to avoid misjudgment caused by relying solely on gradient change. When the echo width ratio is greater than the upper limit of the range interval and multiple defects are identified, the energy concentration is used to screen for real defect points, effectively filtering false local maxima caused by noise. At the same time, an iterative processing mechanism is triggered under specific conditions to dynamically adjust parameters for re-verification, thereby forming a complete defect confirmation process and significantly improving the reliability of the detection results.

[0079] Through the above-mentioned solution, this application effectively reduces the misjudgment rate of minor defects under noise interference, improves the reliability of defect signal verification, and reduces the output of false defect points in multi-defect scenarios, thereby improving the overall performance of the ultrasonic testing system under complex working conditions.

[0080] This application further proposes the following formula for calculating energy concentration:

[0081]

[0082] in, Indicates energy concentration. Indicates the first in the envelope The value of each sampling point, Indicates the peak position. Indicates a fixed window size. This indicates the length of the envelope.

[0083] Specifically, the energy concentration C is a dimensionless index used to quantify the degree of energy concentration in the peak region. It can be achieved by using the ratio of the squares of the envelope values. Normalization is used to eliminate the influence of overall signal amplitude changes, ensuring that the defect confirmation results are not affected by environmental noise fluctuations. The fixed window size K can be understood as a parameter that defines the energy calculation range. It can be achieved by using a preset fixed value or an adaptive adjustment method based on signal characteristics. The purpose is to focus on the core area of ​​the defect response and avoid calculation deviations caused by noise interference at the edge sampling points.

[0084] Specifically, the scheme of the present application forms an energy concentration degree index by calculating the ratio of the sum of the square of the energy of the envelope line within a fixed window around the peak position to the sum of the square of the energy of the entire envelope line. The numerator part is centered on the peak position and the calculation range is limited by the fixed window size K to accurately capture the core energy distribution of the defect response. The denominator part covers the entire envelope line length n to provide a total energy benchmark for normalization processing. This design enables the energy concentration degree to stably represent the degree of energy concentration, effectively eliminating the interference caused by signal amplitude fluctuations, while ensuring that in the detection of small defects, when the energy is highly concentrated, the index can reliably distinguish between real defect signals and scattered noise.

[0085] Through the above scheme, the present application realizes the standardization and accuracy of energy concentration degree calculation, significantly improves the reliability of the defect confirmation process, and effectively reduces the risk of misjudgment and missed detection caused by energy distribution fluctuations in the detection of small defects in mechanical materials.

[0086] In some embodiments of the present application described above, a defect confirmation unit is proposed to confirm the defect result by the energy concentration degree. However, in the implementation process, when the energy concentration degree does not meet the threshold value, there is a lack of explicit iteration processing mechanism to dynamically adjust the envelope line calculation window, which may cause the characteristics of small defects to be masked by the initial window size, resulting in misjudgment.

[0087] To this end, the present application further proposes that the second judgment module further includes an iteration processing unit for performing iteration processing:

[0088] reducing the fixed window for envelope line calculation by a preset fixed ratio, recalculating the envelope line and again extracting the peak amplitude, peak position and echo width, and recalculating the echo width ratio;

[0089] judging whether the recalculated echo width ratio is in a preset range interval, and if so, outputting a typical defect result, otherwise recalculating the gradient change of the peak position and judging whether it is greater than a preset gradient threshold, and if so, outputting a small defect result;

[0090] wherein the iteration processing is terminated when any of the following conditions is met: the number of iterations reaches a preset maximum number of iterations; or the envelope line calculation window is reduced to a preset minimum window value;

[0091] If the iteration is terminated and the defect result is still not output, a no-defect result is finally output.

[0092] The iterative processing unit is a component for dynamically adjusting the envelope line calculation window, and the window size is iteratively optimized to reveal the characteristics of the tiny defects; the preset fixed ratio can be understood as a window reduction ratio factor, which can be set as a constant less than 1, gradually refining the signal analysis without causing the window to be too small; the termination conditions of the iterative processing include the maximum number of iterations and the minimum window value, and the maximum number of iterations can be set to a limited number of iterations, and the minimum window value can be set to the minimum number of sampling points that meet the signal analysis requirements.

[0093] Specifically, the scheme of the present application activates the iterative processing unit when the energy concentration is insufficient, first reduces the envelope line window by a fixed ratio to highlight the signal details, then recalculates the envelope line and extracts the key features, verifies the defect type based on the updated feature values, and if it still does not meet the typical defects, checks the gradient change to identify the tiny defects, and the iterative process is terminated when the maximum number of iterations is reached or the window is too small, thereby ensuring detection accuracy without excessive calculation.

[0094] Through the above scheme, the present application effectively solves the problem that the characteristics of tiny defects are covered by the initial window, improves the accuracy of defect confirmation, and reduces the risk of misjudgment.

[0095] In an ultrasonic testing system, the noise threshold is used to judge the peak amplitude to identify defects. However, a fixed threshold cannot adapt to the dynamic changes of environmental noise, resulting in that tiny defect signals are easily misjudged as noise and missed when the noise level fluctuates, especially when the material environment or detection conditions change, it is difficult for the fixed threshold to accurately capture low-intensity defect signals.

[0096] To this end, the present application further proposes that the noise threshold is determined through a dynamic adjustment mechanism, specifically including: maintaining a peak amplitude queue of historical defect-free samples, the peak amplitude queue containing the peak amplitudes of the last M defect-free samples, where M is a preset positive integer; calculating the mean and standard deviation of all peak amplitudes in the peak amplitude queue, and calculating the noise threshold according to the mean and standard deviation.

[0097] The peak amplitude queue of historical defect-free samples is a dynamic buffer area for storing peak amplitude data of defect-free detection results, which can be implemented using a circular queue or a first-in-first-out queue data structure to ensure that the queue capacity is constant at M and track the background noise intensity changes in real time; the calculation of the mean and standard deviation is a statistical analysis of the queue data, which can be implemented using a sliding window algorithm or an incremental calculation method to efficiently quantify the central tendency and dispersion of noise; the calculation of the noise threshold is to dynamically generate a judgment threshold based on the statistical results, which can be in the form of mean plus k times standard deviation, where k is a preset coefficient, and the purpose is to make the threshold adapt to the noise fluctuation range, avoiding the failure problem of the fixed threshold when the environment changes.

[0098] Specifically, the scheme of the present application continuously collects background noise data by maintaining a peak amplitude queue of historical defect-free samples, ensures that the queue only contains verified defect-free sample peak amplitudes, and thus accurately reflects the current environmental noise level; on this basis, the average value of the queue data represents the typical noise intensity, and the standard deviation represents the noise fluctuation amplitude, both of which together constitute the quantitative indicators of noise characteristics; further, the noise threshold is dynamically generated according to the average value and the standard deviation, so that the threshold can cover the normal noise fluctuation range while remaining sensitive to abnormal signals, thereby accurately distinguishing between small defect signals and background noise when the noise level changes, and avoiding the risk of missed detection.

[0099] Through the above scheme, the present application can dynamically adjust the noise threshold to adapt to the change of environmental noise, thereby effectively identifying low-intensity small defect signals when the material environment or detection conditions fluctuate, and significantly reducing the probability of small defects being misjudged as noise and missed.

[0100] For this, the present application further proposes that the noise threshold is the average value plus twice the standard deviation, and the calculation formula is:

[0101]

[0102] wherein, represents the noise threshold, represents the average value of all peak amplitudes in the peak amplitude queue, represents the standard deviation of all peak amplitudes in the peak amplitude queue.

[0103] Specifically, the scheme of the present application maintains a peak amplitude queue of historical defect-free samples, calculates the average value and standard deviation of the peak amplitudes in the queue in real time, and sets the noise threshold to the average value plus twice the standard deviation. Since under the assumption that noise signals are approximately normally distributed, this threshold covers about 95% of the normal noise data range, so it can effectively distinguish between environmental noise and real defect signals. When the peak amplitude of the newly acquired ultrasonic echo sequence is less than the noise threshold, the system determines that there is no defect; otherwise, further combined with features such as echo width ratio, gradient change, etc. to identify the defect type. This dynamic adjustment mechanism ensures that the threshold can adapt to the noise change in different detection environments, thereby forming a complete noise adaptability processing flow.

[0104] Through the above scheme, the system can more reliably identify small defects in a noisy environment, avoiding the problem of misidentifying noise as defects or missing small defects due to improper threshold setting, thereby improving the accuracy and stability of ultrasonic detection.

[0105] Embodiment two:

[0106] As Figures 2-3As shown, the method for detecting tiny defects inside mechanical material based on ultrasonic wave comprises the following steps:

[0107] Obtaining an ultrasonic echo sequence, the ultrasonic echo sequence comprising amplitude values of a plurality of sampling points;

[0108] Obtaining an envelope line from the ultrasonic echo sequence through moving average processing, the envelope line representing energy distribution of the echo;

[0109] Extracting peak amplitude, peak position and echo width from the envelope line;

[0110] Judging whether the peak amplitude is less than a noise threshold, if yes, outputting a no-defect result, otherwise, calculating an echo width ratio from the echo width and a standard echo width;

[0111] Judging whether the echo width ratio is in a preset range interval, if yes, outputting a typical defect result;

[0112] Otherwise, when the echo width ratio is less than a lower limit of the range interval, calculating gradient change of the peak position from the envelope line; judging whether the gradient change is greater than a preset gradient threshold, if yes, outputting a tiny defect result, otherwise, outputting a no-defect result;

[0113] When the echo width ratio is greater than an upper limit of the range interval, finding all local maximum points from the envelope line, calculating minimum distance between the local maximum points and the peak position; judging whether the minimum distance is less than a preset distance threshold, if yes, outputting a multiple defect result, otherwise, outputting a single large defect result.

[0114] The above content is merely an example and description of the structure of the present application, and those skilled in the art can make various modifications or supplements to the described specific embodiments or replace them with similar ways, as long as they do not deviate from the structure of the present application or exceed the scope defined by the present claims, which shall be within the protection scope of the present application.

[0115] In the description of the present specification, the description of the terms "one embodiment", "example", "specific example" and the like means that the specific features, structures, materials or characteristics described in conjunction with the embodiment or example are included in at least one embodiment or example of the present application. In the present specification, the illustrative description of the above terms does not necessarily refer to the same embodiment or example. Moreover, the described specific features, structures, materials or characteristics can be combined in any one or more embodiments or examples in a suitable manner.

[0116] The preferred embodiments of the application disclosed above are only to facilitate the elucidation of the application. The preferred embodiments do not describe all the details of the application and limit the application to the specific embodiments. Obviously, many modifications and variations can be made in light of the teachings above. The description is chosen and described in order to provide the best illustration of the application and its practical application to those skilled in the art and to enable those skilled in the art to best utilize the application. The application is limited only by the claims and their full scope and equivalents.

Claims

1. A system for detecting small defects in a mechanical material based on ultrasound, characterized in that it comprises: The system comprises: a data acquisition module configured to acquire an ultrasonic echo sequence, the ultrasonic echo sequence comprising amplitude values of a plurality of sampling points; an envelope line generation module configured to obtain an envelope line by moving average processing according to the ultrasonic echo sequence, the envelope line representing an energy distribution of echoes; the envelope line generation process comprises: for each sampling point in the ultrasonic echo sequence, calculating an average value of absolute values of amplitudes of all sampling points in a fixed window centered on the sampling point, and taking the average value as a value of a corresponding sampling point of the envelope line; wherein a size of the fixed window is determined according to a sampling frequency and a center frequency of the ultrasonic wave; the envelope line is calculated by the following formula: wherein represents the value of the i-th sample point in the envelope, represents the amplitude value of the i-th sample point in the ultrasound echo sequence, represents the size of the fixed window, represents the radius of the window, represents a floor operation;​​ a feature extraction module configured to extract a peak amplitude, a peak position and an echo width according to the envelope line; a first judgment module configured to judge whether the peak amplitude is less than a noise threshold, and if yes, output a no-defect result, otherwise, calculate an echo width ratio according to a ratio of the echo width to a standard echo width; a second judgment module configured to judge whether the echo width ratio is in a preset range interval, and if yes, output a typical defect result; otherwise, when the echo width ratio is less than a lower limit of the range interval, calculate a gradient change of the peak position according to the envelope line; judge whether the gradient change is greater than a preset gradient threshold, and if yes, output a micro-defect result, otherwise, output a no-defect result; when the echo width ratio is greater than an upper limit of the range interval, find all local maximum points according to the envelope line, calculate a minimum distance between the local maximum points and the peak position, and judge whether the minimum distance is less than a preset distance threshold, and if yes, output a plurality of defect results, otherwise, output a single large defect result.

2. The system for detecting micro-defects in mechanical material according to claim 1, wherein: the extraction process of the peak amplitude and the peak position comprises: traversing all sampling points in the envelope line, finding a maximum amplitude value as the peak amplitude, and recording an index of the maximum amplitude value in the ultrasonic echo sequence as the peak position; the extraction process of the echo width comprises: taking a preset proportion of the peak amplitude as a width threshold, counting a number of sampling points in the envelope line whose amplitude values continuously exceed the width threshold, and taking the number of sampling points as the echo width.

3. The system for detecting micro-defects in mechanical material according to claim 1, wherein: the process of calculating the gradient change of the peak position according to the envelope line comprises: calculating an absolute value of a difference between two adjacent sampling points of the peak position in the envelope line, and taking the absolute value of the difference as the gradient change; the process of finding all local maximum points according to the envelope line comprises: traversing each sampling point in the envelope line, judging whether a value corresponding to the sampling point in the envelope line is greater than values corresponding to a previous sampling point and a next sampling point of the sampling point in the envelope line, and if yes, taking the sampling point as a local maximum point.

4. The system for detecting small defects in a mechanical material based on ultrasonic waves according to claim 1, characterized in that: The second judgment module further comprises a defect confirmation unit configured to confirm the determined defect result, comprising: Calculate the total energy and peak region energy of the envelope, wherein the peak region energy is calculated with a fixed window centered on the peak position; The energy concentration is calculated by the ratio of the peak region energy to the total energy. When the echo width ratio is less than the lower limit of the range interval and the gradient change is greater than the gradient threshold, it is further determined whether the energy concentration is greater than the preset energy concentration threshold. If yes, the minor defect result is confirmed; otherwise, the defect result is confirmed by performing iterative processing. When the echo width ratio is greater than the upper limit of the range and multiple defect results are determined, only the local maximum point where the energy concentration is greater than the energy concentration threshold is output as the real defect.

5. The system for detecting small defects in the interior of a mechanical material based on ultrasonic waves according to claim 4, characterized in that: The formula for calculating the energy concentration is: wherein, denotes the energy concentration, denotes the value of the sample point in the envelope, denotes the peak position, denotes the fixed window size, denotes the envelope length.

6. The system for detecting small defects in a mechanical material based on ultrasonic waves according to claim 4, characterized in that: The second judgment module further includes an iterative processing unit for performing iterative processing: The envelope calculation window is reduced by a preset fixed ratio, the envelope is recalculated, and the peak amplitude, peak position and echo width are extracted again. The echo width ratio is then recalculated. Determine whether the recalculated echo width ratio is within the preset range. If yes, output typical defect results; otherwise, recalculate the gradient change at the peak position and determine whether it is greater than the gradient threshold. If yes, output minor defect results. The iterative process terminates when any of the following conditions are met: the number of iterations reaches a preset maximum number of iterations; or the envelope calculation window is reduced to a preset minimum window value. If no defective result is output when the iteration terminates, then a defect-free result will be output in the end.

7. The system for detecting small defects in a mechanical material based on ultrasonic waves according to claim 1, characterized in that: The noise threshold is determined through a dynamic adjustment mechanism, specifically including: Maintain a peak amplitude queue of historical defect-free samples, the peak amplitude queue containing the peak amplitudes of the most recent M defect-free samples, where M is a preset positive integer; Calculate the average and standard deviation of all peak amplitudes in the peak amplitude queue, and calculate the noise threshold based on the average and standard deviation.

8. The system for detecting small defects in the interior of a mechanical material based on ultrasonic waves according to claim 7, characterized in that: The noise threshold is the average value plus twice the standard deviation, calculated using the following formula: wherein denotes a noise threshold, denotes the average of all peak amplitudes in the peak amplitude queue, denotes the standard deviation of all peak amplitudes in the peak amplitude queue.

9. Method for detecting small defects in the interior of a mechanical material based on ultrasound, characterized in that, Includes the following steps: Acquire an ultrasonic echo sequence, the ultrasonic echo sequence including amplitude values ​​of multiple sampling points; An envelope representing the energy distribution of the echoes is obtained by moving average processing based on the ultrasonic echo sequence. The envelope generation process includes: for each sampling point in the ultrasonic echo sequence, calculating the average absolute value of the amplitude of all sampling points within a fixed window centered on that sampling point, and using this average value as the value of the corresponding sampling point in the envelope; wherein the size of the fixed window is determined based on the ultrasonic sampling frequency and center frequency; the envelope is calculated using the following formula: wherein represents the value of the i-th sample point in the envelope, represents the amplitude value of the i-th sample point in the ultrasound echo sequence, represents the size of the fixed window, represents the radius of the window, represents a floor operation;​​ The peak amplitude, peak position, and echo width are extracted from the envelope. Determine whether the peak amplitude is less than the noise threshold. If it is, output a defect-free result. Otherwise, calculate the echo width ratio based on the ratio of the echo width to the standard echo width. Determine whether the echo width ratio is within a preset range; if so, output typical defect results. Otherwise, when the echo width ratio is less than the lower limit of the range interval, a gradient change of the peak position is calculated according to the envelope line; whether the gradient change is greater than a preset gradient threshold is judged, yes outputting a micro defect result, otherwise outputting a no defect result; When the echo width ratio is greater than the upper limit of the range interval, all local maximum points are searched according to the envelope line, and the minimum distance between the local maximum points and the peak position is calculated; whether the minimum distance is less than a preset distance threshold is judged, yes outputting multiple defect results, otherwise outputting a single large defect result.

Images