Steel wire rope defect detection method and device, storage medium, equipment, and system

Abstract

The invention discloses a defect detection method and device, a storage medium, equipment and system for a steel wire rope, and relates to the technical field of magnetic flux leakage detection. The invention mainly aims to solve the problems of uneven detection capability of existing magnetic flux leakage nondestructive detection of steel wire rope for defects of different scales and poor environmental adaptability. The invention comprises the following steps: obtaining magnetic flux leakage detection signals of multiple detection channels, and performing signal fusion processing on the multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector; the detection channels are evenly distributed along the circumference of the steel wire rope; performing multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result; adopting an adaptive threshold strategy to perform defect detection on each scale component in the multi-scale decomposition result to obtain an adaptive detection result corresponding to each scale component; performing multi-scale feature fusion processing based on the adaptive detection result to obtain a comprehensive abnormality score, so that defects are classified based on the comprehensive abnormality score.

Description

Technical Field

[0001] The present invention relates to the technical field of magnetic flux leakage detection, and in particular to a method and device, storage medium, equipment and system for detecting defects of a steel wire rope. Background Art

[0002] Wire rope is a critical load-bearing element in many industrial applications, and its integrity directly affects operational safety and efficiency. With the advancement of industrial technology, the demand for accurate and reliable non-destructive testing of wire rope has increased significantly. Among various non-destructive testing (NDT) technologies, magnetic flux leakage (MFL) technology has attracted much attention due to its high sensitivity, non-contact and ability to detect internal defects.

[0003] However, in terms of wire rope defect detection, existing magnetic flux leakage MFL signal processing methods, such as time domain analysis, frequency domain analysis, time-frequency analysis, and machine learning-based methods, have their own limitations when processing complex MFL signals. The detection capabilities of defects of different scales (such as tiny cracks and large-area corrosion) are uneven; the accuracy of identifying weak defect signals in strong background noise is low, and it cannot adapt to different working conditions. Therefore, a wire rope defect detection method is urgently needed to solve the above problems. Summary of the invention

[0004] In view of this, the present invention provides a defect detection method and device, storage medium, equipment, and system for a wire rope, the main purpose of which is to solve the problems of uneven detection capability of existing non-destructive detection of magnetic flux leakage of wire ropes for defects of different scales and poor environmental adaptability.

[0005] According to one aspect of the present invention, a method for detecting defects in a steel wire rope is provided, comprising:

[0006] Acquire magnetic flux leakage detection signals of multiple detection channels, and perform signal fusion processing on the multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector; the detection channels are evenly distributed along the circumference of the wire rope;

[0007] Performing multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result;

[0008] Adopting an adaptive threshold strategy to perform defect detection on each scale component in the multi-scale decomposition result, and obtaining an adaptive detection result corresponding to each scale component;

[0009] A multi-scale feature fusion process is performed based on the adaptive detection result to obtain a comprehensive anomaly score, so that defect classification is performed based on the comprehensive anomaly score.

[0010] Furthermore, the signal fusion processing is performed on the plurality of magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector, including:

[0011] Obtaining component weight factors of the magnetic flux leakage detection signal in radial, tangential and axial directions, respectively, and channel weight factors corresponding to each of the detection channels;

[0012] Based on the component weight factors, first weighted processing is performed on the signal components of each of the magnetic flux leakage detection signals in the radial direction, the tangential direction and the axial direction; and based on the channel weight factors, second weighted processing is performed on the magnetic flux leakage detection signals corresponding to each of the detection channels;

[0013] The signals after the first weighted processing and the second weighted processing are fused to obtain the multi-channel fusion feature vector.

[0014] Furthermore, performing multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result includes:

[0015] Get the preset decomposition scale level;

[0016] Based on the decomposition scale level, the multi-channel fusion feature vector is decomposed level by level using a discrete wavelet transform method to obtain scale components and wavelet coefficients corresponding to each decomposition scale level.

[0017] Furthermore, the method of using an adaptive threshold strategy to perform defect detection on each scale component in the multi-scale decomposition result to obtain an adaptive detection result corresponding to each scale component includes:

[0018] determining an adaptive threshold corresponding to the scale component based on a local mean, a local standard deviation, a local signal-to-noise ratio estimate, and a local probability distribution of the scale component;

[0019] The absolute value of the wavelet coefficient on the scale component is compared with the corresponding adaptive threshold to obtain the adaptive detection result corresponding to the scale component.

[0020] Further, the determining of an adaptive threshold corresponding to the scale component based on a local mean, a local standard deviation, a local signal-to-noise ratio estimate, and a local probability distribution of the scale component includes:

[0021] Calculating a signal-to-noise ratio adjustment factor based on the local signal-to-noise ratio estimate;

[0022] Calculate the difference between the local probability distribution and the reference entropy to obtain an information entropy anomaly index;

[0023] The adaptive threshold corresponding to each of the scale components is obtained by estimating the local mean, the local standard deviation, the signal-to-noise ratio adjustment factor and the information entropy anomaly index using a variational Bayesian method.

[0024] Furthermore, the multi-scale feature fusion processing based on the adaptive detection result is performed to obtain a comprehensive anomaly score, including:

[0025] Based on the adaptive detection results, a Bayesian inference method is used to calculate the abnormal posterior probability corresponding to each scale component;

[0026] Obtain the preset scale weight factor and anomaly scoring function;

[0027] Multi-scale feature fusion processing is performed based on the scale weight factor, the anomaly scoring function and the anomaly posterior probability to obtain the comprehensive anomaly score.

[0028] According to another aspect of the present invention, there is provided a defect detection device for a steel wire rope, comprising:

[0029] A multi-channel fusion module is used to obtain the magnetic flux leakage detection signals of multiple detection channels, and perform signal fusion processing on the multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector; the detection channels are evenly distributed along the circumference of the wire rope;

[0030] A multi-scale decomposition module, used to perform multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result;

[0031] An adaptive detection module, used to perform defect detection on each scale component in the multi-scale decomposition result by adopting an adaptive threshold strategy to obtain an adaptive detection result corresponding to each scale component;

[0032] The anomaly analysis module is used to perform multi-scale feature fusion processing based on the adaptive detection result to obtain a comprehensive anomaly score, so as to classify defects based on the comprehensive anomaly score.

[0033] Furthermore, the multi-channel fusion module is also used for:

[0034] Obtaining component weight factors of the magnetic flux leakage detection signal in radial, tangential and axial directions, respectively, and channel weight factors corresponding to each of the detection channels;

[0035] Based on the component weight factors, first weighted processing is performed on the signal components of each of the magnetic flux leakage detection signals in the radial direction, the tangential direction and the axial direction; and based on the channel weight factors, second weighted processing is performed on the magnetic flux leakage detection signals corresponding to each of the detection channels;

[0036] The signals after the first weighted processing and the second weighted processing are fused to obtain the multi-channel fusion feature vector.

[0037] Furthermore, the multi-scale decomposition module is also used for:

[0038] Get the preset decomposition scale level;

[0039] Based on the decomposition scale level, the multi-channel fusion feature vector is decomposed level by level using a discrete wavelet transform method to obtain scale components and wavelet coefficients corresponding to each decomposition scale level.

[0040] Furthermore, the adaptive detection module includes:

[0041] an adaptive threshold determination unit, configured to determine an adaptive threshold corresponding to the scale component based on a local mean, a local standard deviation, a local signal-to-noise ratio estimate, and a local probability distribution of the scale component;

[0042] The comparison unit is used to compare the absolute value of the wavelet coefficient on the scale component with the corresponding adaptive threshold to obtain the adaptive detection result corresponding to the scale component.

[0043] Furthermore, the adaptive threshold determination unit is further used for:

[0044] Calculating a signal-to-noise ratio adjustment factor based on the local signal-to-noise ratio estimate;

[0045] Calculate the difference between the local probability distribution and the reference entropy to obtain an information entropy anomaly index;

[0046] The adaptive threshold corresponding to each of the scale components is obtained by estimating the local mean, the local standard deviation, the signal-to-noise ratio adjustment factor and the information entropy anomaly index using a variational Bayesian method.

[0047] Furthermore, the abnormality analysis module is also used for:

[0048] Based on the adaptive detection results, a Bayesian inference method is used to calculate the abnormal posterior probability corresponding to each scale component;

[0049] Obtain the preset scale weight factor and anomaly scoring function;

[0050] Multi-scale feature fusion processing is performed based on the scale weight factor, the anomaly scoring function and the anomaly posterior probability to obtain the comprehensive anomaly score.

[0051] According to another aspect of the present invention, there is provided a storage medium, wherein the storage medium stores at least one executable instruction, wherein the executable instruction enables a processor to execute operations corresponding to the above-mentioned wire rope defect detection method.

[0052] According to another aspect of the present invention, there is provided a device, comprising a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus;

[0053] The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the above-mentioned wire rope defect detection method.

[0054] According to yet another aspect of the present invention, there is provided a wire rope defect detection system, comprising a multi-channel magnetic flux leakage sensor array, a processor and a display;

[0055] The multi-channel magnetic flux leakage sensor array is evenly distributed along the circumference of the wire rope and is used to collect leakage magnetic detection signals of multiple detection channels;

[0056] The processor performs operations corresponding to the above-mentioned wire rope defect detection method;

[0057] The display is used to display the comprehensive abnormality score or defect classification result output by the processor.

[0058] By means of the above technical solution, the technical solution provided by the embodiment of the present invention has at least the following advantages:

[0059] The present invention provides a defect detection method and device, storage medium, equipment and system for a steel wire rope. Compared with the prior art, the present invention obtains magnetic flux leakage detection signals of multiple detection channels and performs signal fusion processing on the multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector, which lays the foundation for subsequent multi-scale analysis, ensures the quality and reliability of input data, and improves the environmental adaptability of steel wire rope defect detection. After the multi-scale decomposition result is obtained by performing multi-scale decomposition processing on the multi-channel fusion feature vector, an adaptive threshold strategy is used to perform defect detection on each scale component in the multi-scale decomposition result to obtain an adaptive detection result corresponding to each scale component; then, a multi-scale feature fusion processing is performed based on the adaptive detection result to obtain a comprehensive abnormality score, realizing the analysis and fusion of the multi-channel fusion feature vector at different scales, and combining the detection results of multiple scales, it can comprehensively identify defects of different sizes and types. This multi-scale detection method can simultaneously identify defects of different scales, from tiny cracks to large-area corrosion, greatly improving the comprehensiveness and accuracy of detection.

[0060] The above description is only an overview of the technical solution of the present invention. In order to more clearly understand the technical means of the present invention, it can be implemented according to the contents of the specification. In order to make the above and other purposes, features and advantages of the present invention more obvious and easy to understand, the specific implementation methods of the present invention are listed below. BRIEF DESCRIPTION OF THE DRAWINGS

[0061] Various other advantages and benefits will become apparent to those of ordinary skill in the art by reading the detailed description of the preferred embodiments below. The accompanying drawings are only for the purpose of illustrating the preferred embodiments and are not to be considered as limiting the present invention. Moreover, the same reference symbols are used throughout the accompanying drawings to represent the same components. In the accompanying drawings:

[0062] Figure 1 A schematic flow chart of a method for detecting defects in a steel wire rope provided in an embodiment of the present invention is shown;

[0063] Figure 2 A schematic diagram showing magnetic flux leakage detection signals of multiple detection channels provided by an embodiment of the present invention is shown;

[0064] Figure 3 A schematic diagram of a process of performing signal fusion processing on multiple magnetic leakage detection signals to obtain a multi-channel fusion feature vector in a wire rope defect detection method provided by an embodiment of the present invention is shown;

[0065] Figure 4 A schematic diagram of a process for performing defect detection on each scale component in a multi-scale decomposition result by using an adaptive threshold strategy in a defect detection method for a steel wire rope provided by an embodiment of the present invention to obtain an adaptive detection result corresponding to each scale component;

[0066] Figure 5 A schematic diagram of a process for obtaining a comprehensive abnormality score by performing multi-scale feature fusion processing based on adaptive detection results in a defect detection method for a steel wire rope provided by an embodiment of the present invention is shown;

[0067] Figure 6 A schematic structural diagram of a defect detection device for a steel wire rope provided in an embodiment of the present invention is shown;

[0068] Figure 7 A schematic structural diagram of a device provided by an embodiment of the present invention is shown. DETAILED DESCRIPTION

[0069] The exemplary embodiments of the present disclosure will be described in more detail below with reference to the accompanying drawings. Although the exemplary embodiments of the present disclosure are shown in the accompanying drawings, it should be understood that the present disclosure can be implemented in various forms and should not be limited by the embodiments set forth herein. On the contrary, these embodiments are provided to enable a more thorough understanding of the present disclosure and to fully convey the scope of the present disclosure to those skilled in the art.

[0070] The embodiment of the present invention provides a method for detecting defects in a steel wire rope, such as Figure 1 As shown, the method includes:

[0071] 101. Obtaining magnetic flux leakage detection signals of a plurality of detection channels, and performing signal fusion processing on the plurality of magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector; the detection channels are evenly distributed along the circumference of the steel wire rope;

[0072] In the embodiment of the present invention, the current execution end obtains the leakage magnetic detection signals of multiple detection channels, wherein the multiple detection channels are provided by a multi-channel magnetic flux leakage sensor array, and the multi-channel magnetic flux leakage sensor array is evenly distributed along the circumference of the wire rope, that is, each detection channel is also evenly distributed along the circumference of the wire rope. For example, when the number of detection channels is four, each channel is 90 degrees apart to form a complete annular detection coverage, which is not specifically limited in the embodiment of the present invention. It should be noted that the signal strength detected by each channel is different to a certain extent, such as Figure 2 As shown, the signal strengths of channel 1, channel 2, channel 3, and channel 4 are different.

[0073] In the embodiment of the present invention, after obtaining the magnetic flux leakage detection signals detected by each detection channel, the current execution end also performs signal fusion processing on multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector. Among them, the signal fusion processing represents the merging of the observation data of the magnetic flux leakage sensors on each detection channel, eliminating the noise and uncertainty of the signal, improving the accuracy and reliability of the signal, and also utilizing the complementarity of the magnetic flux leakage sensor array to obtain more comprehensive magnetic flux leakage information. This step lays the foundation for subsequent multi-scale analysis and ensures the quality and reliability of the input data.

[0074] 102. Perform multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result;

[0075] In an embodiment of the present invention, the current execution end performs multi-scale decomposition processing on the multi-channel fusion feature vector obtained in step 101 to obtain a multi-scale decomposition result. Among them, multi-scale decomposition can decompose the multi-channel fusion feature vector into frequency components at multiple scales, and retain the details and overall characteristics of the original multi-channel fusion feature vector, which is based on a combination of wavelet transform and scale transform. Wavelet transform decomposes the multi-channel fusion feature vector into a series of self-contained signals by performing multi-level high-pass and low-pass filtering on the multi-channel fusion feature vector. Scale transform reduces or enlarges the multi-channel fusion feature vector, thereby realizing the analysis of the multi-channel fusion feature vector at different scales. By combining wavelet transform and scale transform, a multi-scale decomposition result can be obtained, that is, the multi-channel fusion feature vector is decomposed into frequency components at multiple scales.

[0076] 103. Perform defect detection on each scale component in the multi-scale decomposition result by using an adaptive threshold strategy to obtain an adaptive detection result corresponding to each scale component;

[0077] In an embodiment of the present invention, the current execution end uses an adaptive threshold strategy to perform defect detection on each scale component in the multi-scale decomposition result obtained in step 102, and obtains an adaptive detection result corresponding to each scale component. Among them, the adaptive threshold is to dynamically adjust the local threshold according to the local characteristics of each scale component. By comparing the signal characteristics with the dynamically calculated local threshold, it can adapt to the signal characteristics of different scales and positions, and improve the accuracy and reliability of detection. Combining the detection results of multiple scales, defects of different sizes and types can be fully identified. This multi-scale detection method can simultaneously identify defects of different scales, from tiny cracks to large-area corrosion, greatly improving the comprehensiveness and accuracy of detection.

[0078] 104. Perform multi-scale feature fusion processing based on the adaptive detection result to obtain a comprehensive anomaly score, so as to perform defect classification based on the comprehensive anomaly score.

[0079] In an embodiment of the present invention, the current execution end performs multi-scale feature fusion processing based on the adaptive detection result obtained in step 103, that is, the features of whether the defect information is contained in each scale component are fused together to obtain a comprehensive anomaly score. The obtained comprehensive anomaly score represents the quantified defect information. The current execution end can perform defect classification processing based on the obtained comprehensive anomaly score, which is not specifically limited in the embodiment of the present invention.

[0080] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to improve the accuracy and reliability of the detection signal, another defect detection method for a wire rope is provided, such as Figure 3 As shown, the step performs signal fusion processing on the plurality of magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector, including:

[0081] 201. Obtain component weight factors of the magnetic flux leakage detection signal in radial, tangential and axial directions, respectively, and channel weight factors corresponding to each of the detection channels;

[0082] In the embodiment of the present invention, since each detection channel is equipped with Hall sensors in radial (r), tangential (θ) and axial (z) directions to capture three-dimensional magnetic field components, the magnetic leakage detection signal of each channel can be expressed as:

[0083]

[0084] in, Indicates i Channel No. t The signal of the sampling point, Indicates the radial i Channel No. t The signal components of the sampling points, Indicates the tangential i Channel No. t The signal components of the sampling points and Indicates the axial i Channel No. t The signal components of the sampling points.

[0085] The current execution end obtains the weight factors of the radial components of the above-mentioned magnetic flux leakage detection signals. , the component weight factor in the tangential direction and the component weight factor in the axial direction ; and obtain the channel weight factors corresponding to each detection channel ; Among them, the number of channels i It is set according to the specific number of on-site detection channels, and the embodiment of the present invention does not make any specific limitation.

[0086] 202. Perform first weighted processing on the signal components of each of the magnetic flux leakage detection signals in radial, tangential and axial directions based on the component weight factors; and perform second weighted processing on the magnetic flux leakage detection signals corresponding to each of the detection channels based on the channel weight factors;

[0087] 203. Fuse the signals after the first weighted processing and the second weighted processing to obtain the multi-channel fusion feature vector.

[0088] In the embodiment of the present invention, the current execution end obtains the component weight factor based on step 201 , , , and the channel weight factor , for each magnetic flux leakage detection signal in the radial direction , the tangential signal component and the axial signal component Perform the first weighted processing to obtain the radial weighted result , weighted results in the tangential direction and the weighted result in the axial direction ; In addition, the current execution end is also based on the channel weight factor The second weighted processing is performed on the magnetic flux leakage detection signals corresponding to each detection channel, and the following formula is obtained:

[0089]

[0090] Next, the current execution end fuses the signals after the first weighted processing and the second weighted processing to obtain the multi-channel fusion feature vector. The specific formula is as follows:

[0091]

[0092] in, F ( t ) represents the multi-channel fusion feature vector.

[0093] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to analyze and process the detection signal from multiple scales on the basis of retaining the details and overall features of the original multi-channel fusion feature vector, another defect detection method for wire rope is provided, the step of performing multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result includes:

[0094] Get the preset decomposition scale level;

[0095] Based on the decomposition scale level, the multi-channel fusion feature vector is decomposed level by level using a discrete wavelet transform method to obtain scale components and wavelet coefficients corresponding to each decomposition scale level.

[0096] In the embodiment of the present invention, the current execution end obtains the preset maximum decomposition scale level, which is recorded as J Next, based on the decomposition scale level, the multi-channel fusion feature vector obtained from step 201 to step 203 is decomposed level by level using discrete wavelet transform (DWT). The specific discrete transformation formula is as follows:

[0097]

[0098] in, The maximum decomposition scale J Next k The coarse-scale approximation coefficients at the positions are For the j Decomposition scale k The detail coefficient at each position, F ( t ) represents the multi-channel fusion feature vector, N Represents the total characteristic signal length. This decomposition method can capture the global trend and local details of the signal at the same time, providing a multi-scale feature representation for subsequent defect detection. The above approximate coefficient and detail factor They can be collectively referred to as wavelet coefficients, which are obtained recursively through the wavelet decomposition process, as shown below:

[0099]

[0100] in, For the j Decomposition scale kThe approximate coefficients at the positions, is the low-pass filter coefficient, is the high-pass filter coefficient, n represents the filter length index, represents the decomposition scale level, Represents the position index, which is used to determine the spatial position of signal sampling and defect detection. j increase, k The range of j Layer k The value range is , N Indicates the total characteristic signal length. is the approximate coefficient of the previous decomposition scale, and the initial condition is the input multi-channel fusion feature vector F ( t ).

[0101] It should be noted that, in the embodiment of the present invention, the orthogonal wavelet basis that best suits the characteristics of the magnetic flux leakage detection signal can be selected by comparing the performance of different wavelet basis functions, and the embodiment of the present invention does not make any specific limitation.

[0102] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to improve the sensitivity of wire rope abnormality detection, another wire rope defect detection is provided, such as Figure 4 As shown, the step adopts an adaptive threshold strategy to perform defect detection on each scale component in the multi-scale decomposition result to obtain an adaptive detection result corresponding to each scale component, including:

[0103] 301. Determine an adaptive threshold corresponding to the scale component based on a local mean, a local standard deviation, a local signal-to-noise ratio estimate, and a local probability distribution of the scale component;

[0104] In the embodiment of the present invention, the current execution end calculates the local mean and local standard deviation of each scale component. The specific formula is as follows:

[0105]

[0106] in, For the j The local mean of the scale components, For the j The local standard deviation of the scale component, is the half width of the sliding window, For the j Decomposition scale k The wavelet coefficients at the positions are t Represents the sampling points of the time series.

[0107] Next, the current execution end estimates the Bayesian adaptive factor based on the given local data. The specific formula is as follows:

[0108]

[0109] in, Indicates j The scale components at the sampling point t The adaptive factor at For the j The integral variable of the adaptive factor on the scale components, Indicates j Given local data of scale components, For a given local data Bayesian adaptive factor The posterior distribution of .

[0110] Then, the signal-to-noise ratio adjustment factor is calculated based on the local signal-to-noise ratio estimation, and the specific formula is as follows:

[0111]

[0112]

[0113] in, For the j The local signal-to-noise ratio estimate of the scale components, For the j The signal-to-noise ratio adjustment factor of the scale component, t Represents the sampling points of the time series.

[0114] Then, by calculating the difference between the local probability distribution and the reference entropy, the abnormal pattern in the signal is effectively captured to obtain the information entropy abnormality index. The specific formula is as follows:

[0115]

[0116]

[0117] in, For the j The information entropy anomaly index of the scale component is t represents the sampling points of the time series, For the k The local probability distribution estimate of the location, is the reference entropy, which can be the average entropy or theoretical entropy of a normal signal. m is the position index within the sliding window, ranging from [ t - , t + ]; For the j The scale components are located in the sliding window m The wavelet coefficients at .

[0118] Finally, the current execution end calculates the local mean of each scale component based on the above calculation , local standard deviation , signal-to-noise ratio adjustment factor and information entropy anomaly index The variational Bayesian method is used to estimate and obtain the adaptive threshold corresponding to each scale component. The specific formula is as follows:

[0119]

[0120] in, For the j The adaptive threshold of the scale components, For the j The Bayesian adaptive factor of the scale components, It is an adjustable sensitivity parameter, usually between 0.5 and 2.

[0121] 302. Compare the absolute value of the wavelet coefficient on the scale component with the corresponding adaptive threshold to obtain the adaptive detection result corresponding to the scale component.

[0122] In the embodiment of the present invention, the current execution end compares the absolute value of the wavelet coefficient on the scale component with the corresponding adaptive threshold to perform defect detection, and obtains the adaptive detection result corresponding to the scale component. The specific formula is as follows:

[0123]

[0124] in, For the j The scale weight k The detection result of the point, 1 means there is a defect, 0 means there is no defect, For the j The scale weight k The wavelet detail coefficient of the point, For the j The scale weight k Adaptive thresholding of points, Indicates j The scale weight kThe absolute value of the wavelet detail coefficient of a point is used to measure the abnormality of the signal at that point. This piecewise function realizes the threshold-based binary detection decision. By comparing the signal characteristics with the dynamically calculated local threshold, it can adapt to the signal characteristics of different scales and positions, and improve the accuracy and reliability of detection. Combining the detection results of multiple scales, defects of different sizes and types can be fully identified. This multi-scale detection method can simultaneously identify defects of different scales, from tiny cracks to large-area corrosion, greatly improving the comprehensiveness and accuracy of detection.

[0125] Further, as a refinement and extension of the specific implementation of the above embodiment, in order to analyze the detection results at various scales, another defect detection method for wire rope is provided, such as Figure 5 As shown, the step of performing multi-scale feature fusion processing based on the adaptive detection result to obtain a comprehensive anomaly score includes:

[0126] 401. Calculate the abnormal posterior probability corresponding to each scale component using the Bayesian inference method based on the adaptive detection result;

[0127] In the embodiment of the present invention, the current execution end uses the Bayesian inference method to calculate the abnormal posterior probability corresponding to each scale component based on the adaptive detection results obtained from step 301 to step 302. The specific formula is as follows:

[0128]

[0129] in, For the The posterior probability of anomaly corresponding to the scale component, Indicated in There are anomalies (defects) in the scale component, Indicates j The detection results of the scale component, is the prior probability, which can be estimated based on historical data. For the j The total probability of the detection result on the scale component, is the likelihood probability, which indicates the probability of detection when defects exist. Indicated in There are no anomalies in the scale component; is the prior probability of no defects, equal to 1- , The probability of false detection when there is no defect.

[0130] 402. Obtaining a preset scale weight factor and anomaly scoring function;

[0131] 403. Perform multi-scale feature fusion processing based on the scale weight factor, the anomaly scoring function, and the anomaly posterior probability to obtain the comprehensive anomaly score.

[0132] In the embodiment of the present invention, the current execution end obtains a preset scale weight factor and the anomaly scoring function h(); then based on the scale weight factor , the anomaly scoring function h() and the anomaly posterior probability obtained in step 401 Multi-scale feature fusion processing is performed to obtain the comprehensive anomaly score. The specific formula is as follows:

[0133]

[0134] in, For sampling point t The comprehensive abnormality score at For the j The scale weight factor on the scale component, is the smoothing threshold function, For the j The scale component at the sampling point t The wavelet coefficients at For sampling point t The local morphological feature vector at contains the geometric and morphological features of the signal at the sampling point t, including defect length, width ratio, number of signal peaks, waveform symmetry ratio, and local magnetic field gradient change rate. is the anomaly scoring function based on the additional feature vector, is the weight coefficient of the additional feature, which is used to balance the contribution of multi-scale detection results and additional features. This fusion method comprehensively considers information at multiple scales and can more accurately assess the severity and credibility of defects.

[0135] It should be noted that, in the embodiment of the present invention, the current execution end can also be based on the comprehensive abnormality score. , classify the detected defects, such as tiny cracks, large-area corrosion, etc. Alternatively, introduce a confidence evaluation mechanism to assign a confidence score to each detection result to assist in subsequent decision-making. Alternatively, morphological opening and closing operations can be used to eliminate isolated false detection points and connect similar detection areas, etc., which are not specifically limited in the embodiments of the present invention.

[0136] The embodiment of the present invention provides a defect detection method for a wire rope. Compared with the prior art, the present invention obtains the leakage magnetic detection signals of multiple detection channels and performs signal fusion processing on the multiple leakage magnetic detection signals to obtain a multi-channel fusion feature vector, which lays the foundation for subsequent multi-scale analysis, ensures the quality and reliability of input data, and improves the environmental adaptability of wire rope defect detection. After the multi-scale decomposition result is obtained by performing multi-scale decomposition processing on the multi-channel fusion feature vector, an adaptive threshold strategy is used to perform defect detection on each scale component in the multi-scale decomposition result to obtain an adaptive detection result corresponding to each scale component; then, based on the adaptive detection result, a multi-scale feature fusion process is performed to obtain a comprehensive abnormality score, realizing the analysis and fusion of the multi-channel fusion feature vector at different scales, and combining the detection results of multiple scales, it can comprehensively identify defects of different sizes and types. This multi-scale detection method can simultaneously identify defects of different scales, from tiny cracks to large-area corrosion, greatly improving the comprehensiveness and accuracy of detection.

[0137] As the above Figure 1 The embodiment of the present invention provides a defect detection device for a steel wire rope, such as Figure 6 As shown, the device comprises:

[0138] A multi-channel fusion module 51 is used to obtain the magnetic flux leakage detection signals of multiple detection channels, and perform signal fusion processing on the multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector; the detection channels are evenly distributed along the circumference of the wire rope;

[0139] A multi-scale decomposition module 52 is used to perform multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result;

[0140] An adaptive detection module 53, configured to perform defect detection on each scale component in the multi-scale decomposition result by adopting an adaptive threshold strategy, and obtain an adaptive detection result corresponding to each scale component;

[0141] The abnormality analysis module 54 is used to perform multi-scale feature fusion processing based on the adaptive detection result to obtain a comprehensive abnormality score, so as to classify defects based on the comprehensive abnormality score.

[0142] Furthermore, the multi-channel fusion module 51 is also used for:

[0143] Obtaining component weight factors of the magnetic flux leakage detection signal in radial, tangential and axial directions, respectively, and channel weight factors corresponding to each of the detection channels;

[0144] Based on the component weight factors, first weighted processing is performed on the signal components of each of the magnetic flux leakage detection signals in the radial direction, the tangential direction and the axial direction; and based on the channel weight factors, second weighted processing is performed on the magnetic flux leakage detection signals corresponding to each of the detection channels;

[0145] The signals after the first weighted processing and the second weighted processing are fused to obtain the multi-channel fusion feature vector.

[0146] Furthermore, the multi-scale decomposition module 52 is also used for:

[0147] Get the preset decomposition scale level;

[0148] Based on the decomposition scale level, the multi-channel fusion feature vector is decomposed level by level using a discrete wavelet transform method to obtain scale components and wavelet coefficients corresponding to each decomposition scale level.

[0149] Furthermore, the adaptive detection module 53 includes:

[0150] an adaptive threshold determination unit, configured to determine an adaptive threshold corresponding to the scale component based on a local mean, a local standard deviation, a local signal-to-noise ratio estimate, and a local probability distribution of the scale component;

[0151] The comparison unit is used to compare the absolute value of the wavelet coefficient on the scale component with the corresponding adaptive threshold to obtain the adaptive detection result corresponding to the scale component.

[0152] Furthermore, the adaptive threshold determination unit is further used for:

[0153] Calculating a signal-to-noise ratio adjustment factor based on the local signal-to-noise ratio estimate;

[0154] Calculate the difference between the local probability distribution and the reference entropy to obtain an information entropy anomaly index;

[0155] The adaptive threshold corresponding to each of the scale components is obtained by estimating the local mean, the local standard deviation, the signal-to-noise ratio adjustment factor and the information entropy anomaly index using a variational Bayesian method.

[0156] Furthermore, the abnormality analysis module 54 is also used for:

[0157] Based on the adaptive detection results, a Bayesian inference method is used to calculate the abnormal posterior probability corresponding to each scale component;

[0158] Obtain the preset scale weight factor and anomaly scoring function;

[0159] Multi-scale feature fusion processing is performed based on the scale weight factor, the anomaly scoring function and the anomaly posterior probability to obtain the comprehensive anomaly score.

[0160] The embodiment of the present invention provides a defect detection device for a wire rope. Compared with the prior art, the present invention obtains the leakage magnetic detection signals of multiple detection channels and performs signal fusion processing on the multiple leakage magnetic detection signals to obtain a multi-channel fusion feature vector, which lays the foundation for subsequent multi-scale analysis, ensures the quality and reliability of input data, and improves the environmental adaptability of wire rope defect detection. After the multi-scale decomposition result is obtained by performing multi-scale decomposition processing on the multi-channel fusion feature vector, an adaptive threshold strategy is used to perform defect detection on each scale component in the multi-scale decomposition result to obtain an adaptive detection result corresponding to each scale component; then, based on the adaptive detection result, a multi-scale feature fusion process is performed to obtain a comprehensive abnormality score, realizing the analysis and fusion of the multi-channel fusion feature vector at different scales, and combining the detection results of multiple scales, it can comprehensively identify defects of different sizes and types. This multi-scale detection method can simultaneously identify defects of different scales, from tiny cracks to large-area corrosion, greatly improving the comprehensiveness and accuracy of detection.

[0161] According to one embodiment of the present invention, a storage medium is provided, wherein the storage medium stores at least one executable instruction, and the computer executable instruction can execute the defect detection method for a steel wire rope in any of the above method embodiments.

[0162] Figure 7 A schematic diagram of the structure of a device provided according to an embodiment of the present invention is shown, and the specific embodiment of the present invention does not limit the specific implementation of the device.

[0163] like Figure 7 As shown, the device may include: a processor, a communication interface, a memory, and a communication bus.

[0164] Wherein: the processor, the communication interface, and the memory communicate with each other via a communication bus.

[0165] Communication interface, used to communicate with other devices such as client or other server network elements.

[0166] The processor is used to execute the program, and specifically can execute the relevant steps of the above-mentioned wire rope defect detection method.

[0167] Specifically, the program may include program codes including computer operation instructions.

[0168] The processor may be a central processing unit (CPU), or an application specific integrated circuit (ASIC), or one or more integrated circuits configured to implement the embodiments of the present invention. The one or more processors included in the device may be processors of the same type, such as one or more CPUs; or processors of different types, such as one or more CPUs and one or more ASICs.

[0169] Memory is used to store programs. The memory may include high-speed RAM memory and may also include non-volatile memory, such as at least one disk storage.

[0170] The program can be specifically used to cause the processor to perform the following operations:

[0171] Acquire magnetic flux leakage detection signals of multiple detection channels, and perform signal fusion processing on the multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector; the detection channels are evenly distributed along the circumference of the wire rope;

[0172] Performing multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result;

[0173] Adopting an adaptive threshold strategy to perform defect detection on each scale component in the multi-scale decomposition result, and obtaining an adaptive detection result corresponding to each scale component;

[0174] A multi-scale feature fusion process is performed based on the adaptive detection result to obtain a comprehensive anomaly score, so that defect classification is performed based on the comprehensive anomaly score.

[0175] According to yet another embodiment of the present invention, there is provided a wire rope defect detection system, comprising a multi-channel magnetic flux leakage sensor array, a processor and a display;

[0176] The multi-channel magnetic flux leakage sensor array is evenly distributed along the circumference of the wire rope and is used to collect leakage magnetic detection signals of multiple detection channels;

[0177] The processor performs operations corresponding to the above-mentioned wire rope defect detection method;

[0178] The display is used to display the comprehensive abnormality score or defect classification result output by the processor.

[0179] Obviously, those skilled in the art should understand that the above modules or steps of the present invention can be implemented by a general computing device, they can be concentrated on a single computing device, or distributed on a network composed of multiple computing devices, and optionally, they can be implemented by a program code executable by a computing device, so that they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be executed in a different order than here, or they can be made into individual integrated circuit modules, or multiple modules or steps therein can be made into a single integrated circuit module for implementation. Thus, the present invention is not limited to any specific combination of hardware and software.

[0180] The above description is only a preferred embodiment of the present invention and is not intended to limit the present invention. For those skilled in the art, the present invention may have various modifications and variations. Any modification, equivalent replacement, improvement, etc. made within the spirit and principle of the present invention shall be included in the protection scope of the present invention.

Claims

1. A method for detecting defects in a steel wire rope, characterized in that: include: Acquire magnetic flux leakage detection signals of multiple detection channels, and perform signal fusion processing on the multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector; The detection channels are evenly distributed along the circumference of the wire rope; Performing multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result; Adopting an adaptive threshold strategy to perform defect detection on each scale component in the multi-scale decomposition result, and obtaining an adaptive detection result corresponding to each scale component; the multi-scale decomposition result includes scale components and wavelet coefficients corresponding to each decomposition scale level; Performing multi-scale feature fusion processing based on the adaptive detection result to obtain a comprehensive anomaly score, so as to classify defects based on the comprehensive anomaly score; The method of using an adaptive threshold strategy to perform defect detection on each scale component in the multi-scale decomposition result to obtain an adaptive detection result corresponding to each scale component includes: determining an adaptive threshold corresponding to the scale component based on a local mean, a local standard deviation, a local signal-to-noise ratio estimate, and a local probability distribution of the scale component; The absolute value of the wavelet coefficient on the scale component is compared with the corresponding adaptive threshold to obtain the adaptive detection result corresponding to the scale component.

2. The method for detecting defects in a steel wire rope according to claim 1, characterized in that: Performing signal fusion processing on the plurality of magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector includes: Obtaining component weight factors of the magnetic flux leakage detection signal in radial, tangential and axial directions, respectively, and channel weight factors corresponding to each of the detection channels; Based on the component weight factors, first weighted processing is performed on the signal components of each of the magnetic flux leakage detection signals in the radial direction, the tangential direction and the axial direction; and based on the channel weight factors, second weighted processing is performed on the magnetic flux leakage detection signals corresponding to each of the detection channels; The signals after the first weighted processing and the second weighted processing are fused to obtain the multi-channel fusion feature vector.

3. The method for detecting defects in a steel wire rope according to claim 1, characterized in that: Performing multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result, including: Get the preset decomposition scale level; Based on the decomposition scale level, the multi-channel fusion feature vector is decomposed level by level using a discrete wavelet transform method to obtain scale components and wavelet coefficients corresponding to each decomposition scale level.

4. The method for detecting defects in a steel wire rope according to claim 1, characterized in that: The step of determining an adaptive threshold corresponding to the scale component based on a local mean, a local standard deviation, a local signal-to-noise ratio estimate, and a local probability distribution of the scale component comprises: Calculating a signal-to-noise ratio adjustment factor based on the local signal-to-noise ratio estimate; Calculate the difference between the local probability distribution and the reference entropy to obtain an information entropy anomaly index; The adaptive threshold corresponding to each of the scale components is obtained by estimating the local mean, the local standard deviation, the signal-to-noise ratio adjustment factor and the information entropy anomaly index using a variational Bayesian method.

5. The method for detecting defects in a steel wire rope according to claim 1, characterized in that: The multi-scale feature fusion processing based on the adaptive detection result to obtain a comprehensive anomaly score includes: Based on the adaptive detection results, a Bayesian inference method is used to calculate the abnormal posterior probability corresponding to each scale component; Obtain the preset scale weight factor and anomaly scoring function; Multi-scale feature fusion processing is performed based on the scale weight factor, the anomaly scoring function and the anomaly posterior probability to obtain the comprehensive anomaly score.

6. A defect detection device for a steel wire rope, used to execute the defect detection method for a steel wire rope according to claim 1, characterized in that: include: A multi-channel fusion module is used to obtain magnetic flux leakage detection signals of multiple detection channels, and perform signal fusion processing on the multiple magnetic flux leakage detection signals to obtain a multi-channel fusion feature vector; The detection channels are evenly distributed along the circumference of the wire rope; A multi-scale decomposition module, used for performing multi-scale decomposition processing on the multi-channel fusion feature vector to obtain a multi-scale decomposition result; the multi-scale decomposition result includes scale components and wavelet coefficients corresponding to each decomposition scale level; An adaptive detection module, used to perform defect detection on each scale component in the multi-scale decomposition result by adopting an adaptive threshold strategy to obtain an adaptive detection result corresponding to each scale component; An anomaly analysis module, used to perform multi-scale feature fusion processing based on the adaptive detection results to obtain a comprehensive anomaly score, so as to classify defects based on the comprehensive anomaly score; The adaptive detection module comprises: an adaptive threshold determination unit, configured to determine an adaptive threshold corresponding to the scale component based on a local mean, a local standard deviation, a local signal-to-noise ratio estimate, and a local probability distribution of the scale component; The comparison unit is used to compare the absolute value of the wavelet coefficient on the scale component with the corresponding adaptive threshold to obtain the adaptive detection result corresponding to the scale component.

7. A storage medium, wherein at least one executable instruction is stored in the storage medium, and the executable instruction executes an operation corresponding to the defect detection method for a steel wire rope as described in any one of claims 1 to 5.

8. A device comprising a processor, a memory, a communication interface and a communication bus, wherein the processor, the memory and the communication interface communicate with each other via the communication bus; The memory is used to store at least one executable instruction, and the executable instruction enables the processor to perform operations corresponding to the defect detection method for a steel wire rope as described in any one of claims 1-5.

9. A wire rope defect detection system, characterized in that: including a multi-channel magnetic flux leakage sensor array, a processor, and a display; The multi-channel magnetic flux leakage sensor array is evenly distributed along the circumference of the wire rope and is used to collect leakage magnetic detection signals of multiple detection channels; The processor executes the defect detection method for the steel wire rope according to any one of claims 1 to 5; The display is used to display the comprehensive abnormality score or defect classification result output by the processor.

Images