Nothing Special   »   [go: up one dir, main page]

WO2021248962A1 - 一种检测及区分钢丝绳内外缺陷的无损检测方法和装置 - Google Patents

一种检测及区分钢丝绳内外缺陷的无损检测方法和装置 Download PDF

Info

Publication number
WO2021248962A1
WO2021248962A1 PCT/CN2021/081320 CN2021081320W WO2021248962A1 WO 2021248962 A1 WO2021248962 A1 WO 2021248962A1 CN 2021081320 W CN2021081320 W CN 2021081320W WO 2021248962 A1 WO2021248962 A1 WO 2021248962A1
Authority
WO
WIPO (PCT)
Prior art keywords
magnetic flux
defect
signal
wire rope
steel wire
Prior art date
Application number
PCT/CN2021/081320
Other languages
English (en)
French (fr)
Inventor
张东来
张恩超
高伟
晏小兰
潘世旻
Original Assignee
哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院)
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) filed Critical 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院)
Priority to EP21758301.2A priority Critical patent/EP3951381B1/en
Priority to US17/436,086 priority patent/US11703477B2/en
Publication of WO2021248962A1 publication Critical patent/WO2021248962A1/zh

Links

Images

Classifications

    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
    • G01N27/72—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables
    • G01N27/82—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws
    • G01N27/83—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields
    • G01N27/85—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields using magnetographic methods
    • G—PHYSICS
    • G01—MEASURING; TESTING
    • G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
    • G01N27/00—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means
    • G01N27/72—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables
    • G01N27/82—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws
    • G01N27/83—Investigating or analysing materials by the use of electric, electrochemical, or magnetic means by investigating magnetic variables for investigating the presence of flaws by investigating stray magnetic fields
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR COUNTING
    • G06N—COMPUTING ARRANGEMENTS BASED ON SPECIFIC COMPUTATIONAL MODELS
    • G06N3/00—Computing arrangements based on biological models
    • G06N3/02—Neural networks

Definitions

  • the invention belongs to the technical field of non-destructive testing, and in particular relates to a non-destructive testing method and device for detecting and distinguishing internal and external defects of a steel wire rope.
  • wire rope As a flexible component, wire rope has strong load capacity, outstanding flexibility and excellent working stability, and is widely used in mining, shipping, construction, transportation and other fields.
  • fatigue damage such as wear, broken wires, and corrosion will inevitably occur.
  • the degree of damage tends to be serious. If the steel wire rope cannot be replaced before the whole rope is broken, it will seriously affect the safety. Production, even threatening equipment and personal safety, causing huge economic losses and adverse social impacts.
  • Defects of wire rope can be divided into external defects and internal defects. As the buried depth of defects increases, the detection of defects will become more and more difficult. Existing detection methods cannot quantitatively detect internal defects, nor can they distinguish between internal and external defects.
  • the electromagnetic detection method is the most effective method at this stage. According to the excitation conditions, it can be divided into saturated excitation and unsaturated excitation. Unsaturated excitation detection has strict requirements on sensors, environment, and methods, and cannot accurately perform quantitative detection and cannot be applied to actual detection. Saturation excitation detection can avoid the above shortcomings, improve the accuracy of quantitative detection, and better apply to actual detection.
  • Saturation excitation detection mainly includes two methods: magnetic flux detection and magnetic flux leakage detection.
  • the magnetic flux detection mainly detects the change of the magnetic flux of the measured object.
  • the magnetic flux includes the main magnetic flux, the leakage magnetic flux and the yoke magnetic flux.
  • the advantages of this method are: the flux value detected is related to the cross-sectional loss area of the tested object; whether the defect is external or internal, magnetic flux nondestructive testing can detect it; but when the axial width of the defect is small, its detection ability Very low, unable to quantitatively detect, and unable to distinguish internal defects.
  • Existing magnetic flux detection methods cannot quantitatively detect all defects, and cannot calculate the buried depth of defects.
  • Magnetic flux leakage detection mainly detects the strength of the leakage magnetic field on the surface of the measured object through a sensor array. Magnetic flux leakage detection has a high recognition rate for defects with small axial width, and the detection of defect width is also more accurate. However, for defects with large axial widths, the information of the defects cannot be accurately identified; the buried depth of the defects seriously affects the detection accuracy, and it is impossible to quantitatively detect all the defects.
  • the existing detection methods cannot distinguish the internal and external defects of the wire rope, and the detection accuracy is very low, and the defect buried depth cannot be calculated.
  • the present invention discloses a non-destructive testing method and device for detecting and distinguishing the internal and external defects of the steel wire rope, which can distinguish the internal and external defects of the steel wire rope, and can also calculate the buried depth of the defects, making the detection more accurate.
  • a non-destructive testing method for detecting and distinguishing internal and external defects of steel wire rope which includes the following steps:
  • Step S10 collecting the magnetic flux signal and the magnetic flux leakage signal of the steel wire rope under test
  • Step S20 preprocessing the magnetic flux signal and magnetic flux leakage signal of the steel wire rope under test
  • Step S30 comparing the preprocessed magnetic flux signal and magnetic leakage signal with the preset magnetic flux signal threshold and the preset magnetic leakage signal threshold, respectively, to calculate the defect location;
  • Step S40 extracting the defect magnetic flux signal and the defect magnetic leakage signal according to the defect position
  • Step S50 Calculate the defect width flw of the wire rope under test according to the defect magnetic flux signal and the defect magnetic flux leakage signal;
  • Step S60 Calculate the defect section loss fs of the tested steel wire rope according to the defect width flw of the tested steel wire rope;
  • Step S70 distinguishing internal and external defects:
  • ⁇ is the preset defect judgment value.
  • ⁇ can be set as the cross-sectional loss rate of a broken wire, or set according to actual conditions.
  • the tested steel wire rope is excited to saturation or near saturation, and then the magnetic flux signal and the magnetic flux leakage signal of the tested steel wire rope are collected.
  • the non-destructive testing method for detecting and distinguishing the internal and external defects of the steel wire rope further includes:
  • Step S80 calculating the buried depth of the defect:
  • f3 is the trained multiple equations or multilayer neural network
  • flw is the defect width of the tested wire rope
  • fs is the defect section loss of the tested wire rope
  • ffs is the virtual section loss.
  • step S70 the trained multi-order equations or multi-layer neural network f2 are obtained by using the following steps:
  • Step721 Design x defect widths, y defect cross-sectional losses, a total of x ⁇ y standard surface defects, x and y are natural numbers;
  • Step722 calculate the peak-to-peak value of the corresponding defect magnetic flux leakage waveform by calculating from step S10 to step S60 for x ⁇ y standard surface defects;
  • Step723 Take the obtained peak-to-peak value of the defect magnetic flux leakage waveform and the defect width of the standard surface defect as input independent variables, and ffs as the output standard quantity, and train to obtain multiple equations or multilayer neural network f2.
  • step S80 the trained multi-order equations or multi-layer neural network f3 are obtained by the following steps:
  • Step821 Design x defect width, y defect section loss, z different buried depths, a total of x ⁇ y ⁇ z standard defects, x, y and z are all natural numbers;
  • Step822 For x ⁇ y ⁇ z standard defects, calculate the corresponding ffs from step S10 to step S70;
  • Step823 Take the obtained ffs, defect cross-section loss and defect width as input independent variables, and the buried depth of the defect as the output standard quantity, and train to obtain multiple equations or multi-layer neural network f3.
  • step S10 includes acquiring the magnetic flux signal of the steel wire rope under test through the magnetic flux detection sensor, acquiring the magnetic flux leakage signal of the steel wire rope under test through the magnetic field intensity detection sensor, and adopting the following steps Preliminary processing of the magnetic flux signal of the steel wire rope to eliminate the influence of the wire rope speed;
  • Y i is the measured flux signal after preliminary treatment of the wire rope
  • S is collected to the measured flux signal wire rope
  • dt represents the time derivative
  • N is the total number of sampling points
  • Y ⁇ Sdl, where Y is the magnetic flux signal of the wire rope under test after preliminary processing, dl represents the differential of the spatial distance, and S is the collected magnetic flux signal of the wire rope under test.
  • the preprocessing of the magnetic flux signal of the steel wire rope under test in step S20 includes outliers removal, noise filtering, baseline elimination, etc., on the magnetic flux signal of the steel wire rope under test, which can improve the signal-to-noise ratio of the magnetic flux signal , More conducive to signal feature extraction.
  • the step of removing outliers on the magnetic flux signal of the steel wire rope under test includes:
  • Y(i) is set as the i-th magnetic flux acquisition signal.
  • the step of performing noise filtering on the magnetic flux signal of the steel wire rope under test includes:
  • Noise filtering is performed on the magnetic flux signal of the steel wire rope under test using adaptive filtering, or wavelet transform, or smoothing filtering, or empirical mode decomposition, wherein noise filtering is performed on the magnetic flux signal of the steel wire rope under test by using smoothing filtering.
  • the calculation formula is:
  • n is the number of data to be averaged
  • N is the total number of sampling points.
  • the step of removing the baseline of the magnetic flux signal of the steel wire rope under test includes:
  • Envelope spectrum extraction, or wavelet decomposition, or window averaging, or empirical mode decomposition is used to remove the baseline of the magnetic flux signal of the steel wire rope under test, wherein the magnetic flux signal of the steel wire rope under test is subjected to empirical mode decomposition
  • the steps for baseline elimination include:
  • preprocessing the magnetic flux leakage signal of the steel wire rope under test includes performing outlier elimination, noise filtering, baseline elimination, and wave noise filtering on each magnetic flux leakage signal of the steel wire rope under test, etc. It can improve the signal-to-noise ratio of the magnetic flux leakage signal, which is more conducive to signal feature extraction.
  • the step of removing outliers on each magnetic flux leakage signal of the steel wire rope under test includes:
  • X i,j is the jth sampled value of the i-th Hall sensor.
  • X i,j is much larger than the value of the front and rear magnetic flux leakage signals:
  • the signal X i,j is obtained after the outliers are eliminated.
  • the steps of performing noise filtering on each magnetic flux leakage signal of the steel wire rope under test include:
  • noise filtering Use adaptive filtering, or wavelet transform, or smoothing filter, or empirical mode decomposition to perform noise filtering on each magnetic leakage signal of the tested steel wire rope; wherein, smooth filtering is used to perform noise filtering on each magnetic leakage signal of the tested steel wire rope.
  • the calculation formula of noise filtering is:
  • n is the number of data to be averaged
  • N is the total number of sampling points
  • k is the number of magnetic field intensity detection sensors that collect the magnetic flux leakage signal of the steel wire rope under test.
  • the step of removing the baseline of each magnetic flux leakage signal of the steel wire rope under test includes:
  • Envelope spectrum extraction, or wavelet decomposition, or window averaging, or empirical mode decomposition is used to eliminate the baseline of each magnetic flux leakage signal of the tested steel wire rope; wherein, the empirical mode decomposition is used to remove the baseline of each magnetic flux leakage signal of the tested steel wire rope.
  • the steps for baseline elimination of magnetic flux leakage signal include:
  • the step of filtering out the wave noise of each magnetic flux leakage signal of the steel wire rope under test includes:
  • Wavelet decomposition, or empirical mode decomposition, or adaptive filtering, or gradient method is used to filter each magnetic flux leakage signal of the steel wire rope under test.
  • the steps of filtering magnetic flux leakage signal for wave noise include:
  • the gradient method is used to realize the first-order differentiation of the image.
  • the gradient at the coordinates (x, y) is represented by a two-dimensional column vector:
  • the modulus of this vector is:
  • the multi-channel magnetic leakage signal is summed to obtain the magnetic leakage sum signal X 2 .
  • calculating the defect location in step S30 includes the following steps:
  • Step S31 setting a preset threshold mp of the magnetic flux signal of the defect of the steel wire rope to be tested, where mp is the peak value of the magnetic flux signal of the smallest defect;
  • Step S32 Compare the magnetic flux signal of the steel wire rope under test with the defect magnetic flux signal preset threshold mp, record multiple sets of continuous magnetic flux signals greater than the defect magnetic flux signal preset threshold mp, and record multiple sets of magnetic flux sampling points
  • Step S34 setting a preset peak-to-peak value vp of the magnetic flux leakage signal of the tested wire rope defect, where vp is the preset peak-to-peak value of the magnetic flux leakage signal of the smallest defect;
  • the magnetic flux leakage signal of the steel wire rope under test is compared with the default peak-to-peak value threshold vp of the defect magnetic flux leakage signal, and multiple sets of continuous magnetic flux leakage signals are recorded at the magnetic leakage sampling points that are greater than the default peak-to-peak value threshold vp of the defect magnetic flux leakage signal, and multiple sets of magnetic flux leakage
  • Step S36 compare the sequence (c1, c2,%) with the sequence (d1, d2,%), if
  • the extraction of the defective magnetic flux signal includes: according to the position information of each group of magnetic flux sampling points, extracting FM points forward and backward from the magnetic flux signal of the wire rope as the defective magnetic flux signal ,
  • FM NO ⁇ SM
  • SM is the number of sampling points of 1 strand pitch
  • NO is a natural number of 5-10;
  • calculating the defect width flw of the steel wire rope under test in step S50 includes the following steps:
  • Step S52 calculating the distance between the maximum value and the minimum value of the defect magnetic flux leakage signal according to the waveform peak point position of the defect magnetic flux leakage signal as the waveform width value Xlw of the defect magnetic flux signal;
  • Step S53 when
  • ⁇ M, the defect width flw (Ylw+Xlw) /2-LF, where LF is the distance between the sensor and the surface of the wire rope.
  • step S60 calculating the defective section loss fs of the steel wire rope under test includes the following steps:
  • step S63 the defect width value flw obtained in step S50 and the peak-to-peak value VPP of the defect magnetic flux signal waveform of S61 are substituted into the multi-order equation set or multilayer neural network in step S62, and the defect section loss of the wire rope is calculated. ⁇ fs.
  • the multi-order equation system or multi-layer neural network f1 trained in step S62 is obtained by adopting the following steps:
  • Step S621 design x defect widths, y defect cross-sectional losses, a total of x ⁇ y standard surface defects, x and y are both natural numbers;
  • Step S622 calculate the corresponding peak-to-peak value of the magnetic flux waveform of the defect through calculation from step S10 to step S60 for x ⁇ y standard surface defects;
  • Step S623 Take the corresponding peak-to-peak value of the magnetic flux waveform of the defect and the defect width of the standard surface defect as input independent variables, and the loss of the defect section as the output standard quantity, and train to obtain a multi-order equation set or multilayer neural network f1.
  • the invention also discloses a non-destructive testing device for detecting and distinguishing the defects inside and outside the steel wire rope, which includes:
  • Excitation structure used to excite the wire rope to saturation or near saturation
  • Magnetic flux detection sensor used to obtain the magnetic flux signal of the steel wire rope under test
  • Magnetic field intensity detection sensor used to obtain the magnetic flux leakage signal of the steel wire rope under test
  • the signal acquisition and processing system includes a signal acquisition unit, a signal preprocessing unit, a defect position calculation unit, a defect signal extraction unit, a defect width calculation unit, a defect section loss calculation unit, an internal and external defect distinguishing unit, and a defect buried depth calculation unit ,in,
  • the signal acquisition unit is used to acquire the magnetic flux signal and the magnetic flux leakage signal of the steel wire rope under test,
  • the signal preprocessing unit is used to preprocess the magnetic flux signal and the magnetic flux leakage signal of the steel wire rope under test
  • the defect position calculation unit is configured to compare the preprocessed magnetic flux signal and magnetic leakage signal with a preset magnetic flux signal threshold and a preset magnetic leakage signal threshold to calculate the defect position;
  • the defect signal extraction unit is used to extract the defect magnetic flux signal and the defect magnetic leakage signal according to the defect position
  • the defect width calculation unit is used to calculate the defect width flw of the steel wire rope under test according to the defect magnetic flux signal and the defect magnetic flux leakage signal;
  • the defect section loss calculation unit is used to calculate the defect section loss fs of the tested steel wire rope according to the defect width flw of the tested steel wire rope;
  • the internal and external defect distinguishing unit is used to distinguish whether it is an internal defect or an external defect according to the defect width flw of the wire rope under test and the peak-to-peak value FV of the defect magnetic flux signal;
  • the defect buried depth calculation unit is used to calculate the defect buried depth according to the defect width flw of the tested steel wire rope, the defect cross-sectional loss fs of the tested steel wire and the virtual cross-sectional loss ffs.
  • the present invention has the following beneficial effects:
  • the steel wire rope is excited to a saturated or nearly saturated state, by detecting and collecting the magnetic flux and the amount of leakage magnetic field of the steel wire rope, through calculation and analysis, not only all types of defects of the steel wire rope can be identified, but also internal and external defects can be distinguished; Not only can the defect be accurately and quantitatively detected, but also the buried depth of the defect can be accurately calculated, with high quantitative accuracy.
  • Figure 1 is a schematic diagram of a non-destructive testing device for detecting and distinguishing internal and external defects of a steel wire rope according to the present invention.
  • Fig. 2 is a schematic diagram of a sensor of a non-destructive inspection device for detecting and distinguishing internal and external defects of a steel wire rope according to the present invention.
  • Fig. 3 is a flow chart of a non-destructive testing method for detecting and distinguishing internal and external defects of a steel wire rope according to the present invention.
  • Fig. 4 is a schematic diagram of a defect magnetic flux signal obtained by an embodiment of the present invention.
  • Fig. 5 is a schematic diagram of a defect magnetic flux leakage signal obtained by an embodiment of the present invention.
  • the reference signs include: 1-excitation structure, 2-sensor, 21-magnetic flux detection sensor, 22-magnetic field intensity detection sensor, 3-signal acquisition and processing system.
  • a non-destructive testing device for detecting and distinguishing internal and external defects of a steel wire rope includes:
  • Excitation structure 1 used to excite the wire rope to saturation or near saturation
  • the sensor 2 as shown in FIG. 2, includes a magnetic flux detection sensor 21 for acquiring the magnetic flux signal of the steel wire rope under test, and a magnetic field intensity detection sensor 22 for acquiring the magnetic flux leakage signal of the steel wire rope under test;
  • a signal acquisition and processing system 3 which includes a signal acquisition unit, a signal preprocessing unit, a defect position calculation unit, a defect signal extraction unit, a defect width calculation unit, a defect section loss calculation unit, and internal and external defects Distinguish the unit and the defect buried depth calculation unit.
  • the signal collection unit is used to collect the magnetic flux signal and the magnetic flux leakage signal of the steel wire rope under test
  • the signal preprocessing unit is used to preprocess the magnetic flux signal and the magnetic flux leakage signal of the steel wire rope under test
  • the defect position calculation unit is configured to compare the preprocessed magnetic flux signal and magnetic leakage signal with a preset magnetic flux signal threshold and a preset magnetic leakage signal threshold to calculate the defect position;
  • the defect signal extraction unit is used to extract the defect magnetic flux signal and the defect magnetic flux leakage signal according to the defect position
  • the defect width calculation unit is used to calculate the defect width flw of the steel wire rope under test according to the defect magnetic flux signal and the defect magnetic flux leakage signal;
  • the defect section loss calculation unit is used to calculate the defect section loss fs of the tested steel wire rope according to the defect width flw of the tested steel wire rope;
  • the internal and external defect distinguishing unit is used to distinguish whether it is an internal defect or an external defect according to the defect width flw of the wire rope under test and the peak-to-peak value FV of the defect magnetic flux signal;
  • the defect burying depth calculation unit is used to calculate the defect burying depth according to the defect width flw of the tested steel wire rope and the defect section loss fs and ffs of the tested steel wire.
  • the signal acquisition and processing system adopts the following non-destructive testing method for detecting and distinguishing the internal and external defects of the steel wire rope for processing.
  • the non-destructive testing method for detecting and distinguishing internal and external defects of a steel wire rope includes the following steps:
  • Step10 Collect the detection signal of the wire rope under test
  • Step20 Pre-processing the detection signal of the steel wire rope under test
  • Step30 Calculate the defect location
  • Step40 Extract the defect signal
  • Step50 Calculate the defect width
  • Step60 Calculate the loss rate of the defect section
  • Step70 Distinguish internal and external defects
  • Step80 Calculate the buried depth of defects.
  • the detection signal of the steel wire rope under test collected in step 10 includes the magnetic flux signal and the magnetic flux leakage signal, which specifically include:
  • the magnetic flux signal of the steel wire rope under test is obtained by the magnetic flux detection sensor, and the magnetic flux leakage signal of the steel wire rope under test is obtained by the magnetic field intensity detection sensor; since the magnetic flux signal is affected by the speed of the steel wire rope, the speed of the steel wire rope cannot be detected accurately and in real time. So it is necessary to eliminate the influence of wire rope speed.
  • the magnetic flux signal S is processed through the integration of time by an integrator, and then the data is collected through equal-space sampling or the magnetic flux signal S is collected through the equal-distance integration of the space through the integrator, as shown in the following formula:
  • dt represents the differentiation of time
  • N is the total number of sampling points
  • dl represents the differentiation of space distance
  • the preprocessing of the measured wire rope detection signal includes the preprocessing of the magnetic flux signal and the preprocessing of the magnetic flux leakage signal.
  • the preprocessing of the magnetic flux signal includes outlier elimination, noise filtering, baseline elimination, etc., which can improve the magnetic flux signal.
  • the signal-to-noise ratio is more conducive to signal feature extraction.
  • Step21 Eliminate outliers on the magnetic flux signal Y, and set Y(i) as the i-th magnetic flux acquisition signal.
  • Step22 Use adaptive filtering, or wavelet transform, or smoothing filtering, or empirical mode decomposition to perform noise filtering on the magnetic flux signal of the tested steel wire rope, where smooth filtering is used to perform noise filtering on the magnetic flux signal of the tested steel wire rope
  • the calculation formula of noise filtering is:
  • n is the number of data to be averaged, and N is the total number of sampling points;
  • Step23 Baseline elimination of the above signals.
  • the methods used for baseline elimination include but are not limited to envelope spectrum extraction, wavelet decomposition, window averaging, empirical mode decomposition, etc.
  • the following is an empirical mode decomposition method: find the above signals All the maximum points and minimum points of the data sequence Y 1 (i) are fitted to the upper and lower envelopes of the original sequence with a cubic spline function; the mean value of the upper and lower envelopes is m1; Subtract m1 from the data sequence to get a new sequence Y 2 (i) minus the low frequency, that is, Y 2 (i) Y 1 (i)-m1.
  • n is the number of data to be averaged, and N is the total number of sampling points;
  • the preprocessing of the magnetic flux leakage signal is mainly to eliminate outliers, noise filtering, baseline elimination, and wave filtering for each magnetic flux leakage signal, which can improve the signal-to-noise ratio of the magnetic flux leakage signal and is more conducive to signal feature extraction.
  • the following are the specific steps:
  • Step24 Eliminate outliers for each channel of magnetic flux leakage signal X, and set X i,j as the jth sampled value of the i-th Hall sensor.
  • X i,j is much larger than the front and back magnetic flux leakage signal values:
  • Step25 Perform noise filtering on each of the above signals.
  • the methods used for noise filtering include but are not limited to adaptive filtering, wavelet transform, smoothing filtering, empirical mode decomposition, etc.
  • the following is a smoothing filtering processing method:
  • n is the number of data to be averaged
  • N is the total number of sampling points
  • k is the number of sensor channels
  • Step27 Perform wave noise filtering on each channel of magnetic flux leakage signal of the tested steel wire rope, and adopt wavelet decomposition, or empirical mode decomposition, or adaptive filtering, or gradient method to each channel of magnetic flux leakage of the tested steel wire rope The signal is filtered out of wave noise, wherein the steps of using the gradient method to filter out the wave noise of each channel of magnetic flux leakage signal of the steel wire rope under test include:
  • the gradient method is used to realize the first-order differentiation of the image.
  • the gradient at the coordinates (x, y) is represented by a two-dimensional column vector:
  • the modulus of this vector is:
  • the multi-channel magnetic leakage signal is summed to obtain the magnetic leakage sum signal X 2 .
  • step30 The method of calculating the defect location described in step30 is as follows:
  • Step31 Set the preset threshold mp for the magnetic flux signal of the defect of the steel wire rope to be tested, where mp is the peak value of the magnetic flux signal of the smallest defect;
  • Step32 Compare the magnetic flux signal of the wire rope under test with the preset threshold mp, and record multiple sets of continuous magnetic flux sampling points.
  • Step34 Set the default peak-to-peak value vp of the magnetic flux leakage signal of the tested wire rope, where vp is the default peak-to-peak value of the magnetic flux leakage signal of the smallest defect;
  • Step34 Compare the magnetic flux signal of the steel wire rope under test with the preset threshold value vp, and record multiple sets of continuous magnetic flux leakage sampling points.
  • Step36 Compare (c1, c2, %) and (d1, d2, 7), if there is a relatively close value, that is
  • the extraction of the defect signal in step 40 includes the extraction of the magnetic flux signal of the defect and the extraction of the magnetic leakage signal of the defect, which specifically include:
  • step50 the method of calculating defect width is as follows:
  • Step52 According to the position of the peak point of the defect magnetic flux leakage signal waveform, calculate the distance between the maximum value and the minimum value of the defect magnetic flux leakage signal as the waveform width value Xlw of the defect magnetic flux signal;
  • Step53 When
  • ⁇ M, the defect width flw (Ylw+Xlw)/2-LF, LF Is the distance between the sensor and the surface of the wire rope.
  • step60 the method of calculating the loss of defect section is as follows:
  • Step63 Substitute the defect width value flw obtained in step S50 and the peak-to-peak value VPP of the defect magnetic flux signal of S61 into the multiple equations or multilayer neural network in step S62 to calculate the accurate defect metal section loss ⁇ fs.
  • the multi-level equations or multi-layer neural network f1 trained in step62 is obtained by the following method:
  • Step621 Design x width flw, y metal cross-section loss area fs, a total of x ⁇ y standard surface defects, x and y are natural numbers;
  • Step622 calculate the peak-to-peak value VPP of the corresponding defect magnetic flux waveform through step S10 to step S60 for x ⁇ y standard injuries;
  • Step623 Take waveform peak-to-peak value VPP and standard injury waveform width value flw as input independent variables, and cross-sectional loss area fs as output standard quantity, train to obtain multiple equations or multilayer neural network f1.
  • step 70 the method for distinguishing internal and external defects includes the following steps:
  • Step73 Substitute the defect width value flw obtained in step S50 and the peak-to-peak value FV of the defect magnetic flux leakage signal of S71 into the multiple equations or multilayer neural network in step S72 to calculate the virtual section loss ffs;
  • Step74 Compare the fs in ffs and S63, if
  • the multi-level equations or multi-layer neural network f2 trained in step72 is obtained by the following method:
  • Step721 Design x width flw, y metal cross-section loss area fs, a total of x ⁇ y standard surface defects, x and y are natural numbers;
  • Step722 The peak-to-peak value FV of the corresponding defect magnetic flux leakage waveform is obtained by calculating from step S10 to step S60 for x ⁇ y standard injuries;
  • Step723 Take the peak-to-peak value FV of the waveform and the standard injury waveform width value flw as input independent variables, and ffs as the output standard quantity, and train to obtain multiple equations or multilayer neural network f2.
  • step80 the method for calculating the buried depth of defects is as follows:
  • Step81 According to the result of S70, if the defect is external, the buried depth of the defect is 0, otherwise the buried depth of the defect is not 0, proceed to step S82;
  • Step83 Substitute the defect width value flw obtained in S50, the fs obtained in S60 and the ffs obtained in S70 into the multi-order equation set or multilayer neural network in step S82, and the buried depth fd of the defect is calculated.
  • step82 The method of training multiple equations or multi-layer neural network f3 of step82 is as follows:
  • Step821 Design x width flw, y metal cross-section loss area fs, z different buried depth fd, a total of x ⁇ y ⁇ z standard defects, x, y and z are all natural numbers;
  • Step822 calculate the corresponding ffs from steps S10 to S70 for x ⁇ y ⁇ z standard injuries
  • Step823 Use ffs, fs, and flw as input independent variables, and fd as the output standard quantity, and train to obtain multi-order equations or multi-layer neural network f3.
  • the wire rope is excited to saturation or nearly saturation, the magnetic flux and magnetic leakage signal of the wire rope are collected and preprocessed, the defect magnetic signal is extracted, and the magnetic flux signal and magnetic leakage signal of the defect are respectively calculated and analyzed ,
  • the fusion analysis of the calculation results of the two signals can distinguish internal and external defects, and further quantitatively calculate the defects, and the final result of the defect buried depth obtained is accurate.

Landscapes

  • Chemical & Material Sciences (AREA)
  • Physics & Mathematics (AREA)
  • Health & Medical Sciences (AREA)
  • Life Sciences & Earth Sciences (AREA)
  • General Physics & Mathematics (AREA)
  • General Health & Medical Sciences (AREA)
  • Engineering & Computer Science (AREA)
  • Biochemistry (AREA)
  • Analytical Chemistry (AREA)
  • Electrochemistry (AREA)
  • Immunology (AREA)
  • Chemical Kinetics & Catalysis (AREA)
  • Pathology (AREA)
  • Theoretical Computer Science (AREA)
  • Biophysics (AREA)
  • Mathematical Physics (AREA)
  • Software Systems (AREA)
  • General Engineering & Computer Science (AREA)
  • Computing Systems (AREA)
  • Molecular Biology (AREA)
  • Evolutionary Computation (AREA)
  • Data Mining & Analysis (AREA)
  • Computational Linguistics (AREA)
  • Biomedical Technology (AREA)
  • Artificial Intelligence (AREA)
  • Investigating Or Analyzing Materials By The Use Of Magnetic Means (AREA)

Abstract

一种检测及区分钢丝绳内外缺陷的无损检测方法和装置，该方法包括以下步骤：采集被测钢丝绳的磁通信号和漏磁信号；对被测钢丝绳的磁通信号和漏磁信号进行预处理；根据预处理后的磁通信号、漏磁信号分别与预设磁通信号阈值、预设漏磁信号阈值进行对比，计算缺陷位置；根据缺陷位置，提取缺陷磁通信号和缺陷漏磁信号；根据缺陷磁通信号和缺陷漏磁信号，计算被测钢丝绳的缺陷宽度flw；根据被测钢丝绳的缺陷宽度flw，计算被测钢丝绳的缺陷截面损失量fs；计算并区分内外部缺陷。本方法不仅可以识别钢丝绳的所有类型缺陷，而且可以区分内外部缺陷，还可以对缺陷的进行精确定量检测，准确计算缺陷的埋藏深度。

Description

一种检测及区分钢丝绳内外缺陷的无损检测方法和装置 技术领域
本发明属于无损检测技术领域，尤其涉及一种检测及区分钢丝绳内外缺陷的无损检测方法和装置。
背景技术
钢丝绳作为一种挠性部件具有强劲的载重能力、突出的柔韧性和卓越的工作稳定性，在矿山、航运、建筑、运输等领域得到广泛应用。然而，钢丝绳在长期使用过程中，必然会产生磨损、断丝、锈蚀等疲劳损伤，随着使用周期的延长，损伤程度也趋于严重，如不能在整绳破断之前替换钢丝绳，将严重影响安全生产，甚至威胁设备及人身安全，造成巨大经济损失及不良的社会影响。钢丝绳的缺陷可以分为外部缺陷和内部缺陷。随着缺陷埋藏深度的增大，缺陷的检出将变得越来越困难。现有的检测方法无法定量检测内部缺陷，也无法区分内外部缺陷。
电磁检测法是现阶段最有效的方法，根据励磁条件可分为饱和励磁和非饱和励磁。非饱和励磁检测对传感器、环境和方式等都有较严格的要求，而且无法准确地进行定量检测，不能应用于实际检测。饱和励磁检测则可以避免以上缺点，提高定量检测的精度，更好地应用于实际检测。
饱和励磁检测主要有磁通检测和漏磁检测两种方法。磁通检测主要检测被测物体的磁通变化量，磁通包括主磁通、漏磁通和磁轭磁通等。此方法的优点是：其检测的通量值与被测对象的截面损失面积相关；无论缺陷在外部还是内部，磁通无损检测都能检测；但是当缺陷的轴向宽度较小时，其检测能力非常低，无法定量检测，而且无法区分内部缺陷。现有的磁通检测方法无法定量检测所有缺陷，且无法计算缺陷埋藏深度。漏磁检测主要通过传感器阵列检测被测物体表面的漏磁场强度。漏磁检测对轴向宽度较小的缺陷有很高识别率，对缺陷宽度的检测也较为准确。但是对于轴向宽度较大的缺陷，无法准确识别缺陷的信息；缺陷埋藏深度严重影响检测精度，无法定量检测所有缺陷。
因此现有的检测方法无法对钢丝绳内外部缺陷进行区分，而且检测精度很低，无法计算缺陷埋藏深度。
发明内容
针对以上技术问题，本发明公开了一种检测及区分钢丝绳内外缺陷的无损检测方法和装置，可以对钢丝绳内外部缺陷进行区分，还可以计算得到缺陷埋藏深度，使得检测更加精准。
对此，本发明采用的技术方案为：
一种检测及区分钢丝绳内外缺陷的无损检测方法，其包括以下步骤：
步骤S10，采集被测钢丝绳的磁通信号和漏磁信号；
步骤S20，对被测钢丝绳的磁通信号和漏磁信号进行预处理；
步骤S30，根据预处理后的磁通信号、漏磁信号分别与预设磁通信号阈值、预设漏磁信号阈值进行对比，计算缺陷位置；
步骤S40，根据缺陷位置，提取缺陷磁通信号和缺陷漏磁信号；
步骤S50，根据缺陷磁通信号和缺陷漏磁信号，计算被测钢丝绳的缺陷宽度flw；
步骤S60，根据被测钢丝绳的缺陷宽度flw，计算被测钢丝绳的缺陷截面损失量fs；
步骤S70，区分内外部缺陷：
采用公式FV＝FH–FL计算所述缺陷漏磁信号的波形峰峰值FV，其中，FH为所述缺陷漏磁信号的波形峰值，FL为所述缺陷漏磁信号的波形谷值；
设计关系函数ffs＝f2(FV,flw)，其中，f2为训练的多次方程组或者多层神经网络，ffs为虚拟截面损失量；
将步骤S50中得到的缺陷宽度flw和所述缺陷漏磁信号的波形峰峰值FV，代入所述设计关系函数中，计算得到ffs；
对比ffs和fs，如果|fs-ffs|＞μ，此缺陷为内部缺陷，否则为外部缺陷；
其中，μ为预设的缺陷判定值。
进一步的，μ可设置为一根断丝的截面损失率，或者根据实际情况设置。
进一步的，将被测钢丝绳励磁至饱和或近似饱和状态，然后采集被测钢丝绳的磁通信号和漏磁信号。
作为本发明的进一步改进，所述的检测及区分钢丝绳内外缺陷的无损检测方法还包括：
步骤S80，计算缺陷埋藏深度：
根据S70结果，如果缺陷为外部缺陷，则缺陷的埋藏深度为0，否则缺陷的埋藏深度不为0，则进行以下步骤；
根据以下公式计算缺陷的埋藏深度fd，
fd＝f3(fs,ffs,flw)，
其中，f3为训练的多次方程组或者多层神经网络，flw为被测钢丝绳的缺陷宽度，fs为被测钢丝绳的缺陷截面损失量，ffs为虚拟截面损失量。
作为本发明的进一步改进，步骤S70中，训练的多次方程组或者多层神经网络f2采用以下步骤得到：
Step721：设计x个缺陷宽度，y个缺陷截面损失量，总共x×y个标准表面缺陷，x和y均为自然数；
Step722：对x×y个标准表面缺陷经过步骤S10至步骤S60的计算得到相应的缺陷漏磁波形峰峰值；
Step723：将得到的缺陷漏磁波形峰峰值和标准表面缺陷的缺陷宽度作为输入自变量，ffs作为输出标准量，训练得到多次方程组或者多层神经网络f2。
作为本发明的进一步改进，步骤S80中，训练的多次方程组或者多层神经网络f3采用以下步骤得到：
Step821：设计x个缺陷宽度，y个缺陷截面损失量，z个不同埋藏深度，总共x×y×z个标准缺陷，x、y和z均为自然数；
Step822：对x×y×z个标准缺陷经过步骤S10至步骤S70的计算得到相应的ffs；
Step823：将得到的ffs、缺陷截面损失量和缺陷宽度作为输入自变量，缺陷的埋藏深度作为输出标准量，训练得到多次方程组或者多层神经网络f3。
作为本发明的进一步改进，步骤S10中包括，通过磁通检测传感器获取被测钢丝绳的磁通信号，通过磁场强度检测传感器获取所述被测钢丝绳的漏磁信号，并采用以下步骤对获取的被测钢丝绳的磁通信号进行初步处理，消除钢丝绳速度的影响；
对磁通信号采用如下公式通过积分器对时间的积分处理，然后通过等空间采样采集数据，
Figure PCTCN2021081320-appb-000001
其中Y i为初步处理后的被测钢丝绳的磁通信号，S为采集到的被测钢丝绳的磁通信号，dt表示对时间的微分，N为总采样点数；
或者采用以下公式对磁通信号通过积分器的对空间进行等距离积分处理采集；
Y＝∫Sdl，其中Y为初步处理后的被测钢丝绳的磁通信号，dl表示对空间距离的微分，S为采集到的被测钢丝绳的磁通信号。
作为本发明的进一步改进，步骤S20中对被测钢丝绳的磁通信号进行预处理包括对被测钢丝绳的磁通信号进行野点剔除、噪声滤波、基线消除等，能够提高磁通信号的信噪比，更利于信号的特征提取。
作为本发明的进一步改进，对所述被测钢丝绳的磁通信号进行野点剔除的步骤包括：
对所述被测钢丝绳的磁通信号Y进行野点剔除，设Y(i)为第i个磁通采集信号，当Y(i)远大于前后磁通信号值时，Y(i)＝[Y(i-1)+Y(i+1)]/2(i＝1,2,…,N)，野点剔除处理之后得到信号Y 1(i)，N为总采样点数。
作为本发明的进一步改进，对所述被测钢丝绳的磁通信号进行噪声滤波的步骤包括：
采用自适应滤波、或者小波变换、或者平滑滤波、或者经验模态分解对所述被测钢丝绳的磁通信号进行噪声滤波，其中，采用平滑滤波对所述被测钢丝绳的磁通信号进行噪声滤波的计算公式为：
Figure PCTCN2021081320-appb-000002
其中，式中n为求均值的数据个数，N为总采样点数。
作为本发明的进一步改进，对所述被测钢丝绳的磁通信号进行基线消除的步骤包括：
采用包络谱提取、或小波分解、或窗口平均、或经验模态分解对所述被测钢丝绳的磁通信号进行基线消除，其中，采用经验模态分解对所述被测钢丝绳的磁通信号进行基线消除的步骤包括：
找出信号数据序列Y 2(i)的所有极大值点和极小值点，将其用三次样条函数分别拟合为原序列的上包络线和下包络线；上包络线和下包络线的均值为m1；将数据序列Y 2(i)减去m1得到一个减去低频的新序列Y 3(i)，即Y 3(i)＝Y 2(i)－m1。
作为本发明的进一步改进，步骤S20中，对被测钢丝绳的漏磁信号进行预处理包 括对被测钢丝绳的每路漏磁信号进行野点剔除、噪声滤波、基线消除、股波噪声滤除等，能够提高漏磁信号的信噪比，更利于信号的特征提取。
作为本发明的进一步改进，对所述被测钢丝绳的每路漏磁信号进行野点剔除的步骤包括：
对每路漏磁信号X进行野点剔除，设X i,j为第i个霍尔传感器的第j个采样值，当X i,j远大于前后漏磁信号值时：
Figure PCTCN2021081320-appb-000003
野点剔除处理之后得到信号X i,j。
作为本发明的进一步改进，对所述被测钢丝绳的每路漏磁信号进行噪声滤波的步骤包括：
采用自适应滤波、或者小波变换、或者平滑滤波、或者经验模态分解对被测钢丝绳的每路漏磁信号进行噪声滤波；其中，采用平滑滤波对所述被测钢丝绳的每路漏磁信号进行噪声滤波的计算公式为：
Figure PCTCN2021081320-appb-000004
其中，式中n为求均值的数据个数，N为总采样点数，k为采集被测钢丝绳的漏磁信号的磁场强度检测传感器路数。
作为本发明的进一步改进，对所述被测钢丝绳的每路漏磁信号进行基线消除的步骤包括：
采用包络谱提取、或者小波分解、或者窗口平均、或者经验模态分解对所述被测钢丝绳的每路漏磁信号进行基线消除；其中，采用经验模态分解对所述被测钢丝绳的每路漏磁信号进行基线消除的步骤包括：
找出野点剔除处理后的漏磁信号数据序列Xy的所有极大值点和极小值点，将其用三次样条函数分别拟合为原序列的上包络线和下包络线，上包络线和下包络线的均值为m1；将原数据序列减去n1得到减去低频的新序列X1，即X1＝Xy－n1。
作为本发明的进一步改进，对所述被测钢丝绳的每路漏磁信号进行股波噪声滤除的步骤包括：
采用小波分解、或者经验模态分解、或者自适应滤波、或者梯度法对所述被测钢 丝绳的每路漏磁信号进行股波噪声滤除，其中，采用梯度法对所述被测钢丝绳的每路漏磁信号进行股波噪声滤除的步骤包括：
采用梯度法实现图像的一阶微分，对于图像X 1(x,y)，其在坐标(x,y)处的梯度是二维列向量表示：
Figure PCTCN2021081320-appb-000005
这个向量的模是：
Figure PCTCN2021081320-appb-000006
对多路漏磁信号进行求和处理，得到漏磁和信号X 2。
作为本发明的进一步改进，步骤S30中计算缺陷位置包括以下步骤：
步骤S31，设置被测钢丝绳的缺陷磁通信号预设阈值mp，mp为最小缺陷的磁通信号峰值；
步骤S32，将被测钢丝绳的磁通信号与缺陷磁通信号预设阈值mp比较，记录多组连续的磁通信号大于缺陷磁通信号预设阈值mp的磁通采样点，多组磁通采集点的轴向坐标为[c11,c12……c1a]，[c21,c22……c2b]，……；
步骤S33，计算每组磁通采集点的轴向坐标的平均值，c1＝(c11+c12+……+c1a)/a，c2＝(c21+c22+……+c2b)/b，……，得到序列(c1，c2，……)；
步骤S34，设置被测钢丝绳的缺陷漏磁信号预设峰峰值阈值vp，vp为最小缺陷的漏磁信号预设峰峰值；
将被测钢丝绳的漏磁信号与缺陷漏磁信号预设峰峰值阈值vp比较，记录多组连续的漏磁信号大于缺陷漏磁信号预设峰峰值阈值vp的漏磁采样点，多组漏磁采样点的轴向坐标为[d11,d12……d1e]，[d21,d22……d2f]，……，得到序列(d1，d2，……)；
步骤S35，计算每组漏磁采集点的轴向坐标最大最小值的平均值，即d1＝(d11+d1e)/2，d2＝(d21+d2f)/2，……；
步骤S36，对比序列(c1，c2，……)和序列(d1，d2，……)，如果|ci-dj|<M，其中M 为钢丝绳股距，则保留ci，舍弃dj，否则ci和dj都保留，计算结果则为缺陷位置。
作为本发明的进一步改进，步骤S40，所述提取缺陷磁通信号包括：根据每组磁通采样点的位置信息，对钢丝绳的磁通信号向前和向后提取FM个点作为缺陷磁通信号，其中FM＝NO×SM，SM为1个股距的采样点数量，NO为5-10的自然数；
所述提取缺陷漏磁信号包括：根据每组漏磁采集点的位置信息，对钢丝绳的漏磁信号向前和向后提取LFM个点作为缺陷漏磁信号，LSM为1个股距的采样点数量，LFM＝LNO×LSM，LNO为5-10的自然数。
作为本发明的进一步改进，步骤S50中计算被测钢丝绳的缺陷宽度flw包括以下步骤：
步骤S51，先根据公式h(s)＝df(s)/ds(s＝1,2,…,k)求解被测钢丝绳的缺陷磁通信号的微分结果h(s)，其中，k为缺陷磁通信号的数据个数，f(s)为缺陷磁通信号的数据；然后根据缺陷磁通信号的波形峰值点位置，向前取h(s)的最大值的位置，向后取h(s)的最小值的位置，计算最大值和最小值之间的间距作为所述缺陷磁通信号的波形宽度值Ylw；
步骤S52，根据缺陷漏磁信号的波形峰值点位置，计算缺陷漏磁信号的最大值和最小值之间的间距作为所述缺陷磁通信号的波形宽度值Xlw；
步骤S53，当|Ylw-Xlw|＜M，M为钢丝绳股距，选取Ylw、Xlw中的较大值为缺陷宽度flw；当|Ylw-Xlw|≥M，缺陷宽度flw＝(Ylw+Xlw)/2-LF，其中LF为传感器距离钢丝绳表面的距离。
作为本发明的进一步改进，步骤S60，计算被测钢丝绳的缺陷截面损失量fs包括如下步骤：
步骤S61，将得到的缺陷磁通信号的波形峰值T和波形基线值L，通过公式VPP＝|T–L|计算所述缺陷磁通信号的波形峰峰值VPP；
步骤S62，设计关系函数fs＝f1(VPP,flw)，其中，f1为训练的多次方程组或者多层神经网络；
步骤S63，将步骤S50中得到的缺陷宽度值flw和S61的缺陷磁通信号的波形峰峰值VPP，代入到步骤S62中的多次方程组或者多层神经网络中，计算得到钢丝绳的缺陷截面损失量fs。
作为本发明的进一步改进，步骤S62中所述训练的多次方程组或者多层神经网络 f1采用以下步骤得到：
步骤S621，设计x个缺陷宽度，y个缺陷截面损失量，总共x×y个标准表面缺陷，x和y均为自然数；
步骤S622：对x×y个标准表面缺陷经过步骤S10至步骤S60的计算得到相应的缺陷磁通波形峰峰值；
步骤S623：将相应的缺陷磁通波形峰峰值和标准表面缺陷的缺陷宽度作为输入自变量，缺陷截面损失量作为输出标准量，训练得到多次方程组或者多层神经网络f1。
本发明还公开了一种检测及区分钢丝绳内外缺陷的无损检测装置，其包括：
励磁结构，用于将钢丝绳励磁至饱和或近似饱和状态；
磁通检测传感器，用于获取被测钢丝绳的磁通信号；
磁场强度检测传感器，用于获取所述被测钢丝绳的漏磁信号；
以及信号采集和处理系统，其采用如上所述的检测及区分钢丝绳内外缺陷的无损检测方法进行处理；
所述信号采集和处理系统包括信号采集单元、信号预处理单元、缺陷位置计算单元、缺陷信号提取单元、缺陷宽度计算单元、缺陷截面损失量计算单元、内外部缺陷区分单元和缺陷埋藏深度计算单元，其中，
所述信号采集单元用于采集被测钢丝绳的磁通信号和漏磁信号，
所述信号预处理单元用于对被测钢丝绳的磁通信号和漏磁信号进行预处理；
所述缺陷位置计算单元用于根据预处理后的磁通信号、漏磁信号分别与预设磁通信号阈值、预设漏磁信号阈值进行对比，计算缺陷位置；
所述缺陷信号提取单元用于根据缺陷位置，提取缺陷磁通信号和缺陷漏磁信号；
所述缺陷宽度计算单元用于根据缺陷磁通信号和缺陷漏磁信号，计算被测钢丝绳的缺陷宽度flw；
所述缺陷截面损失量计算单元用于根据被测钢丝绳的缺陷宽度flw，计算被测钢丝绳的缺陷截面损失量fs；
所述内外部缺陷区分单元用于根据被测钢丝绳的缺陷宽度flw和缺陷磁通信号的波形峰峰值FV，区分是内部缺陷还是外部缺陷；
所述缺陷埋藏深度计算单元用于根据被测钢丝绳的缺陷宽度flw、被测钢丝绳的缺陷截面损失量fs和虚拟截面损失量ffs计算缺陷埋藏深度。
与现有技术相比，本发明的有益效果为：
采用本发明的技术方案，将钢丝绳励磁至饱和或近似饱和状态，通过检测并采集钢丝绳的磁通量和漏磁场量，通过计算和分析，不仅可以识别钢丝绳的所有类型缺陷，而且可以区分内外部缺陷；不仅可以对缺陷的进行精确定量检测，而且可以准确计算缺陷的埋藏深度，具有较高的定量精度。
附图说明
图1是本发明一种检测及区分钢丝绳内外缺陷的无损检测装置的示意图。
图2是本发明一种检测及区分钢丝绳内外缺陷的无损检测装置的传感器的示意图。
图3是本发明一种检测及区分钢丝绳内外缺陷的无损检测方法的流程图。
图4是本发明实施例得到的缺陷磁通信号的示意图。
图5是本发明实施例得到的缺陷漏磁信号的示意图。
附图标记包括：1-励磁结构，2-传感器，21-磁通检测传感器，22-磁场强度检测传感器，3-信号采集和处理系统。
具体实施方式
下面对本发明的较优的实施例作进一步的详细说明。
如图1所示，一种检测及区分钢丝绳内外缺陷的无损检测装置，其包括：
励磁结构1，用于将钢丝绳励磁至饱和或近似饱和状态；
传感器2，如图2所示，其包括用于获取被测钢丝绳的磁通信号的磁通检测传感器21、和用于获取所述被测钢丝绳的漏磁信号的磁场强度检测传感器22；
以及信号采集和处理系统3，所述信号采集和处理系统包括信号采集单元、信号预处理单元、缺陷位置计算单元、缺陷信号提取单元、缺陷宽度计算单元、缺陷截面损失量计算单元、内外部缺陷区分单元和缺陷埋藏深度计算单元。
其中，所述信号采集单元用于采集被测钢丝绳的磁通信号和漏磁信号，
所述信号预处理单元用于对被测钢丝绳的磁通信号和漏磁信号进行预处理；
所述缺陷位置计算单元用于根据预处理后的磁通信号、漏磁信号分别与预设磁通信号阈值、预设漏磁信号阈值进行对比，计算缺陷位置；
所述缺陷信号提取单元用于根据缺陷位置，提取缺陷磁通信号和缺陷漏磁信号；
所述缺陷宽度计算单元用于根据缺陷磁通信号和缺陷漏磁信号，计算被测钢丝绳的缺陷宽度flw；
所述缺陷截面损失量计算单元用于根据被测钢丝绳的缺陷宽度flw，计算被测钢丝绳的缺陷截面损失量fs；
所述内外部缺陷区分单元用于根据被测钢丝绳的缺陷宽度flw和缺陷磁通信号的波形峰峰值FV，区分是内部缺陷还是外部缺陷；
所述缺陷埋藏深度计算单元用于根据被测钢丝绳的缺陷宽度flw、被测钢丝绳的缺陷截面损失量fs和ffs计算缺陷埋藏深度。
所述信号采集和处理系统采用如下所述的检测及区分钢丝绳内外缺陷的无损检测方法进行处理。
具体而言，如图3所示，所述的检测及区分钢丝绳内外缺陷的无损检测方法包括以下步骤：
Step10：采集被测钢丝绳检测信号；
Step20：对被测钢丝绳检测信号进行预处理；
Step30：计算缺陷位置；
Step40：提取缺陷信号；
Step50：计算缺陷宽度；
Step60：计算缺陷截面损失率；
Step70：区分内外部缺陷；
Step80：计算缺陷埋藏深度。
其中，step10中采集被测钢丝绳检测信号包括磁通信号和漏磁信号，具体包括：
通过磁通检测传感器获取被测钢丝绳的磁通信号，通过磁场强度检测传感器获取所述被测钢丝绳的漏磁信号；由于磁通信号受到钢丝绳的速度影响，且无法准确实时的检测钢丝绳的速度，所以需要消除钢丝绳速度的影响。对磁通信号S通过积分器对时间的积分处理，然后通过等空间采样采集数据或者对磁通信号S通过积分器的对空间进行等距离积分处理采集，如下公式：
Figure PCTCN2021081320-appb-000007
或者Y＝∫Sdl
其中dt表示对时间的微分，N为总采样点数，dl表示对空间距离的微分。
step20中，对被测钢丝绳检测信号进行预处理包括磁通信号的预处理和漏磁信号的预处理，磁通信号的预处理包括野点剔除、噪声滤波、基线消除等，能够提高磁通信号的信噪比，更利于信号的特征提取。以下是具体的步骤：
Step21：对磁通信号Y进行野点剔除，设Y(i)为第i个磁通采集信号，当Y(i)远大于前后磁通信号值时，Y(i)＝[Y(i-1)+Y(i+1)]/2(i＝1,2,…,N)，野点剔除处理之后得到信号Y 1(i)，N为总采样点数；
Step22：采用自适应滤波、或者小波变换、或者平滑滤波、或者经验模态分解对所述被测钢丝绳的磁通信号进行噪声滤波，其中，采用平滑滤波对所述被测钢丝绳的磁通信号进行噪声滤波的计算公式为：
Figure PCTCN2021081320-appb-000008
其中，式中n为求均值的数据个数，N为总采样点数；
Step23：对以上信号进行基线消除，基线消除采用的方法包括但不限于包络谱提取、小波分解、窗口平均、经验模态分解等，以下是一种经验模态分解的方法：找出以上信号数据序列Y 1(i)的所有极大值点和极小值点，将其用三次样条函数分别拟合为原序列的上和下包络线；上下包络线的均值为m1；将数据序列减去m1可得到一个减去低频的新序列Y 2(i)，即Y 2(i)＝Y 1(i)－m1。式中n为求均值的数据个数，N为总采样点数；
漏磁信号的预处理主要是对每路漏磁信号进行野点剔除、噪声滤波、基线消除、股波滤除等，能够提高漏磁信号的信噪比，更利于信号的特征提取。以下是具体的步骤：
Step24：对每路漏磁信号X进行野点剔除，设X i,j为第i个霍尔传感器的第j个采样值，当X i,j远大于前后漏磁信号值时：
Figure PCTCN2021081320-appb-000009
野点剔除处理之后得到信号X i,j；
Step25：对以上每路信号进行噪声滤除，噪声滤除采用的方法包括但不限于自适应 滤波、小波变换、平滑滤波、经验模态分解等，以下是一种平滑滤波的处理方法：
Figure PCTCN2021081320-appb-000010
式中n为求均值的数据个数，N为总采样点数，k为传感器路数；
Step26：对所述被测钢丝绳的每路漏磁信号进行基线消除，采用包络谱提取、或者小波分解、或者窗口平均、或者经验模态分解对所述被测钢丝绳的每路漏磁信号进行基线消除；其中，采用经验模态分解对所述被测钢丝绳的每路漏磁信号进行基线消除的步骤包括：找出野点剔除处理后的漏磁信号数据序列Xy的所有极大值点和极小值点，将其用三次样条函数分别拟合为原序列的上包络线和下包络线，上包络线和下包络线的均值为n1；将原数据序列减去n1得到减去低频的新序列X 1，即X1＝Xy－n1。
Step27：对所述被测钢丝绳的每路漏磁信号进行股波噪声滤除，采用小波分解、或者经验模态分解、或者自适应滤波、或者梯度法对所述被测钢丝绳的每路漏磁信号进行股波噪声滤除，其中，采用梯度法对所述被测钢丝绳的每路漏磁信号进行股波噪声滤除的步骤包括：
采用梯度法实现图像的一阶微分，对于图像X 1(x,y)，其在坐标(x,y)处的梯度是二维列向量表示：
Figure PCTCN2021081320-appb-000011
这个向量的模是：
Figure PCTCN2021081320-appb-000012
对多路漏磁信号进行求和处理，得到漏磁和信号X 2。
step30中所述计算缺陷位置的方法如下：
Step31：设置被测钢丝绳的缺陷磁通信号预设阈值mp，mp为最小缺陷的磁通信号峰值；
Step32：被测钢丝绳的磁通信号与预设阈值mp比较，记录多组连续的磁通采样点，多组采集点的轴向坐标为[c11,c12……c1a]，[c21,c22……c2b]，……；
Step33：计算每组磁通采集点的轴向坐标的平均值，c1＝(c11+c12+……+c1a)/a，c2＝(c21+c22+……+c2b)/b，……；
Step34：设置被测钢丝绳的缺陷漏磁信号预设峰峰值阈值vp，vp为最小缺陷的漏磁信号预设峰峰值；
Step34：被测钢丝绳的磁通信号与预设阈值vp比较，记录多组连续的漏磁采样点，多组采集点的轴向坐标为[d11,d12……d1e]，[d21,d22……d2f]，……；
Step35：计算每组漏磁采集点的轴向坐标最大最小值的平均值，即d1＝(d11+d1e)/2，d2＝(d21+d2f)/2，……；
Step36：对比(c1，c2，……)和(d1，d2，……)，如果有比较接近的数值，即|ci-dj|<M，M为钢丝绳股距，则保留ci，舍弃dj，否则ci和dj都保留，计算结果则为缺陷位置。
step40中所述提取缺陷信号包括提取缺陷的磁通信号和提取缺陷的漏磁信号，具体包括：
Step41：根据每组磁通采集点的位置信息，对钢丝绳的磁通信号向前和向后提取FM个点，SM为1个股距的采样点数量，FM＝NO×SM，NO为5-10，也可以根据实际检测情况设置，截取的数据作为缺陷磁通信号；
Step42：根据每组漏磁采集点的位置信息，对钢丝绳的漏磁信号向前和向后提取LFM个点，LSM为1个股距的采样点数量，LFM＝LNO×LSM，LNO为5-10，也可以根据实际检测情况设置，截取的数据作为缺陷漏磁信号。
step50中，所述计算缺陷宽度的方法如下：
Step51：先通过公式h(s)＝df(s)/ds(s＝1,2,…,k)求解所述缺陷磁通信号的微分结果h(s)，其中，k缺陷磁通信号的数据个数，f(s)为缺陷磁通信号的数据；然后根据缺陷磁通信号的波形峰值点位置，向前取h(s)的最大值的位置，向后取h(s)的最小值的位置，计算最大值和最小值之间的间距作为所述缺陷磁通信号的波形宽度值Ylw，如图4所示；
Step52：根据缺陷漏磁信号波形峰值点位置，计算缺陷漏磁信号的最大值和最小值 之间的间距作为所述缺陷磁通信号的波形宽度值Xlw；
Step53：当|Ylw-Xlw|＜M，M为钢丝绳股距，选取较大值为缺陷宽度flw；当|Ylw-Xlw|≥M，缺陷宽度flw＝(Ylw+Xlw)/2-LF，LF为传感器距离钢丝绳表面的距离。
step60中，所述计算缺陷截面损失量的方法如下：
Step61：得到所述缺陷磁通信号的波形峰值T和波形基线值L，通过公式VPP＝|T–L|计算所述缺陷磁通信号的波形峰峰值VPP；
Step62：设计关系函数fs＝f1(VPP,flw)，其中，f1为训练的多次方程组或者多层神经网络；
Step63：将步骤S50中得到的缺陷宽度值flw和S61的缺陷磁通信号的波形峰峰值VPP，代入到步骤S62中的多次方程组或者多层神经网络中，计算得到准确的缺陷金属截面损失量fs。
step62中所述训练的多次方程组或者多层神经网络f1采用如下方法得到：
Step621：设计x个宽度flw，y个金属截面损失面积fs，总共x×y个标准表面缺陷，x和y均为自然数；
Step622：对x×y个标准伤经过步骤S10至步骤S60的计算得到相应的缺陷磁通波形峰峰值VPP；
Step623：将波形峰峰值VPP和标准伤波形宽度值flw作为输入自变量，截面损失面积fs作为输出标准量，训练得到多次方程组或者多层神经网络f1。
step70中，所述区分内外部缺陷方法包括如下步骤：
Step71：得到所述缺陷漏磁信号的波形峰值FH和波形谷值FL，如图5所示，通过公式FV＝FH–FL计算所述缺陷漏磁信号的波形峰峰值FV；
Step72：设计关系函数ffs＝f2(FV,flw)，其中，f2为训练的多次方程组或者多层神经网络；
Step73：将步骤S50中得到的缺陷宽度值flw和S71的缺陷漏磁信号的波形峰峰值FV，代入到步骤S72中的多次方程组或者多层神经网络中，计算得到虚拟截面损失量ffs；
Step74：对比ffs和S63中的fs，如果|fs-ffs|＞μ，其中μ可设置为一根断丝的截面 损失率，或者根据实际情况设置，此缺陷为内部缺陷，否则为外部缺陷。
其中step72所述训练的多次方程组或者多层神经网络f2采用如下方法得到：
Step721：设计x个宽度flw，y个金属截面损失面积fs，总共x×y个标准表面缺陷，x和y均为自然数；
Step722：对x×y个标准伤经过步骤S10至步骤S60的计算得到相应的缺陷漏磁波形峰峰值FV；
Step723：将波形峰峰值FV和标准伤波形宽度值flw作为输入自变量，ffs作为输出标准量，训练得到多次方程组或者多层神经网络f2。
step80中，所述计算缺陷埋藏深度的方法如下：
Step81：根据S70结果，如果缺陷为外部，则缺陷的埋藏深度为0，否则缺陷的埋藏深度不为0，进行S82步骤；
Step82：设计关系函数fd＝f3(fs,ffs,flw)，其中，f3为训练的多次方程组或者多层神经网络；
Step83：将S50中得到的缺陷宽度值flw，S60得到的fs和S70得到的ffs代入到步骤S82中的多次方程组或者多层神经网络中，计算得到缺陷的埋藏深度fd。
其中step82的训练多次方程组或者多层神经网络f3方法如下：
Step821：设计x个宽度flw，y个金属截面损失面积fs，z个不同埋藏深度fd，总共x×y×z个标准缺陷，x、y和z均为自然数；
Step822：对x×y×z个标准伤经过步骤S10至步骤S70的计算得到相应的ffs；
Step823：将ffs，fs和flw作为输入自变量，fd作为输出标准量，训练得到多次方程组或者多层神经网络f3。
采用本实施例的技术方案，将钢丝绳励磁至饱和或近似饱和，对钢丝绳的磁通量和漏磁信号进行采集和预处理，提取缺陷磁信号，对缺陷的磁通信号和漏磁信号分别进行计算分析，对两种信号的计算结果进行融合分析，可以区分内外部缺陷，进一步对缺陷进行定量计算，最后得到的缺陷埋藏深度结果准确。
以上内容是结合具体的优选实施方式对本发明所作的进一步详细说明，不能认定本发明的具体实施只局限于这些说明。对于本发明所属技术领域的普通技术人员来说，在不脱离本发明构思的前提下，还可以做出若干简单推演或替换，都应当视为属于本 发明的保护范围。

Claims (10)

  1. 一种检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于，其包括以下步骤：
    步骤S10，采集被测钢丝绳的磁通信号和漏磁信号；
    步骤S20，对被测钢丝绳的磁通信号和漏磁信号进行预处理；
    步骤S30，根据预处理后的磁通信号、漏磁信号分别与预设磁通信号阈值、预设漏磁信号阈值进行对比，计算缺陷位置；
    步骤S40，根据缺陷位置，提取缺陷磁通信号和缺陷漏磁信号；
    步骤S50，根据缺陷磁通信号和缺陷漏磁信号，计算被测钢丝绳的缺陷宽度flw；
    步骤S60，根据被测钢丝绳的缺陷宽度flw，计算被测钢丝绳的缺陷截面损失量fs；
    步骤S70，区分内外部缺陷：
    采用公式FV＝FH–FL计算所述缺陷漏磁信号的波形峰峰值FV，其中，FH为所述缺陷漏磁信号的波形峰值，FL为所述缺陷漏磁信号的波形谷值；
    设计关系函数ffs＝f2(FV,flw)，其中，f2为训练的多次方程组或者多层神经网络，ffs为虚拟截面损失量；
    将步骤S50中得到的缺陷宽度flw和所述缺陷漏磁信号的波形峰峰值FV，代入所述设计关系函数中，计算得到虚拟截面损失量ffs；
    对比ffs和fs，如果|fs-ffs|＞μ，此缺陷为内部缺陷，否则为外部缺陷；
    其中，μ为预设的缺陷判定值。
  2. 根据权利要求1所述的检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于：还包括：
    步骤S80，计算缺陷埋藏深度：
    根据S70结果，如果缺陷为外部缺陷，则缺陷的埋藏深度为0，否则缺陷的埋藏深度不 为0，则进行以下步骤；
    根据以下公式计算缺陷的埋藏深度fd，
    fd＝f3(fs,ffs,flw)，
    其中，f3为训练的多次方程组或者多层神经网络，flw为被测钢丝绳的缺陷宽度，fs为被测钢丝绳的缺陷截面损失量，ffs为虚拟截面损失量。
  3. 根据权利要求2所述的检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于：
    步骤S70中，训练的多次方程组或者多层神经网络f2采用以下步骤得到：
    Step721：设计x个缺陷宽度，y个缺陷截面损失量，总共x×y个标准表面缺陷，x和y均为自然数；
    Step722：对x×y个标准表面缺陷经过步骤S10至步骤S60的计算得到相应的缺陷漏磁波形峰峰值；
    Step723：将得到的缺陷漏磁波形峰峰值和标准表面缺陷的缺陷宽度作为输入自变量，ffs作为输出标准量，训练得到多次方程组或者多层神经网络f2；
    步骤S80中，训练的多次方程组或者多层神经网络f3采用以下步骤得到：
    Step821：设计x个缺陷宽度，y个缺陷截面损失量，z个不同埋藏深度，总共x×y×z个标准缺陷，x、y和z均为自然数；
    Step822：对x×y×z个标准缺陷经过步骤S10至步骤S70的计算得到相应的ffs；
    Step823：将得到的ffs、缺陷截面损失量和缺陷宽度作为输入自变量，缺陷的埋藏深度作为输出标准量，训练得到多次方程组或者多层神经网络f3。
  4. 根据权利要求2所述的检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于：
    步骤S10中包括，通过磁通检测传感器获取被测钢丝绳的磁通信号，通过磁场强度检 测传感器获取所述被测钢丝绳的漏磁信号，并采用以下步骤对获取的被测钢丝绳的磁通信号进行初步处理，消除钢丝绳速度的影响；
    对磁通信号采用如下公式通过积分器对时间的积分处理，然后通过等空间采样采集数据，
    Figure PCTCN2021081320-appb-100001
    其中Y i为初步处理后的被测钢丝绳的磁通信号，S为采集到的被测钢丝绳的磁通信号，dt表示对时间的微分，N为总采样点数；
    或者采用以下公式对磁通信号通过积分器的对空间进行等距离积分处理采集；
    Y＝∫Sdl，其中Y为初步处理后的被测钢丝绳的磁通信号，dl表示对空间距离的微分，S为采集到的被测钢丝绳的磁通信号。
  5. 根据权利要求4所述的检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于：步骤S20中对被测钢丝绳的磁通信号进行预处理包括对被测钢丝绳的磁通信号进行野点剔除、噪声滤波、基线消除；
    其中，对所述被测钢丝绳的磁通信号进行野点剔除的步骤包括：
    对所述被测钢丝绳的磁通信号Y进行野点剔除，设Y(i)为第i个磁通采集信号，当Y(i)远大于前后磁通信号值时，Y(i)＝[Y(i-1)+Y(i+1)]/2(i＝1,2,…,N)，野点剔除处理之后得到信号Y 1(i)，N为总采样点数；
    对所述被测钢丝绳的磁通信号进行噪声滤波的步骤包括：
    采用自适应滤波、或者小波变换、或者平滑滤波、或者经验模态分解对所述被测钢丝绳的磁通信号进行噪声滤波，其中，采用平滑滤波对所述被测钢丝绳的磁通信号进行噪声滤波的计算公式为：
    Figure PCTCN2021081320-appb-100002
    其中，式中n为求均值的数据个数，N为总采样点数；
    对所述被测钢丝绳的磁通信号进行基线消除的步骤包括：
    采用包络谱提取、或小波分解、或窗口平均、或经验模态分解对所述被测钢丝绳的磁通信号进行基线消除，其中，采用经验模态分解对所述被测钢丝绳的磁通信号进行基线消除的步骤包括：
    找出信号数据序列Y 2(i)的所有极大值点和极小值点，将其用三次样条函数分别拟合为原序列的上包络线和下包络线；上包络线和下包络线的均值为m1；将数据序列Y 2(i)减去m1得到一个减去低频的新序列Y 3(i)，即Y 3(i)＝Y 2(i)－m1；
    步骤S20中，对被测钢丝绳的漏磁信号进行预处理包括对被测钢丝绳的每路漏磁信号进行野点剔除、噪声滤波、基线消除、股波噪声滤除；
    其中，对所述被测钢丝绳的每路漏磁信号进行野点剔除的步骤包括：
    对每路漏磁信号X进行野点剔除，设X i,j为第i个霍尔传感器的第j个采样值，当X i,j远大于前后漏磁信号值时：
    Figure PCTCN2021081320-appb-100003
    野点剔除处理之后得到信号X i,j；
    对所述被测钢丝绳的每路漏磁信号进行噪声滤波的步骤包括：
    采用自适应滤波、或者小波变换、或者平滑滤波、或者经验模态分解对被测钢丝绳的每路漏磁信号进行噪声滤波；其中，采用平滑滤波对所述被测钢丝绳的每路漏磁信号进行噪声滤波的计算公式为：
    Figure PCTCN2021081320-appb-100004
    其中，式中n为求均值的数据个数，N为总采样点数，k为采集被测钢丝绳的漏磁信号的磁场强度检测传感器路数；
    对所述被测钢丝绳的每路漏磁信号进行基线消除的步骤包括：
    采用包络谱提取、或者小波分解、或者窗口平均、或者经验模态分解对所述被测钢丝绳的每路漏磁信号进行基线消除；其中，采用经验模态分解对所述被测钢丝绳的每路漏磁信号进行基线消除的步骤包括：
    找出野点剔除处理后的漏磁信号数据序列Xy的所有极大值点和极小值点，将其用三次样条函数分别拟合为原序列的上包络线和下包络线，上包络线和下包络线的均值为n1；将原数据序列减去n1得到减去低频的新序列X1，即X1＝Xy－n1；
    对所述被测钢丝绳的每路漏磁信号进行股波噪声滤除的步骤包括：
    采用小波分解、或者经验模态分解、或者自适应滤波、或者梯度法对所述被测钢丝绳的每路漏磁信号进行股波噪声滤除，其中，采用梯度法对所述被测钢丝绳的每路漏磁信号进行股波噪声滤除的步骤包括：
    采用梯度法实现图像的一阶微分，对于图像X 1(x,y)，其在坐标(x,y)处的梯度是二维列向量表示：
    Figure PCTCN2021081320-appb-100005
    这个向量的模是：
    Figure PCTCN2021081320-appb-100006
    对多路漏磁信号进行求和处理，得到漏磁和信号X 2。
  6. 根据权利要求5所述的检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于：
    步骤S30中计算缺陷位置包括以下步骤：
    步骤S31，设置被测钢丝绳的缺陷磁通信号预设阈值mp，mp为最小缺陷的磁通信号峰值；
    步骤S32，将被测钢丝绳的磁通信号与缺陷磁通信号预设阈值mp比较，记录多组连续的磁通信号大于缺陷磁通信号预设阈值mp的磁通采样点，多组磁通采集点的轴向坐标为[c11,c12……c1a]，[c21,c22……c2b]，……；
    步骤S33，计算每组磁通采集点的轴向坐标的平均值，c1＝(c11+c12+……+c1a)/a，c2＝(c21+c22+……+c2b)/b，……，得到序列(c1，c2，……)；
    步骤S34，设置被测钢丝绳的缺陷漏磁信号预设峰峰值阈值vp，vp为最小缺陷的漏磁信号预设峰峰值；
    将被测钢丝绳的漏磁信号与缺陷漏磁信号预设峰峰值阈值vp比较，记录多组连续的漏磁信号大于缺陷漏磁信号预设峰峰值阈值vp的漏磁采样点，多组漏磁采样点的轴向坐标为[d11,d12……d1e]，[d21,d22……d2f]，……，得到序列(d1，d2，……)；
    步骤S35，计算每组漏磁采集点的轴向坐标最大最小值的平均值，即d1＝(d11+d1e)/2，d2＝(d21+d2f)/2，……；
    步骤S36，对比序列(c1，c2，……)和序列(d1，d2，……)，如果|ci-dj|<M，其中M为钢丝绳股距，则保留ci，舍弃dj，否则ci和dj都保留，计算结果则为缺陷位置。
  7. 根据权利要求6所述的检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于：
    步骤S40，所述提取缺陷磁通信号包括：根据每组磁通采样点的位置信息，对钢丝绳的 磁通信号向前和向后提取FM个点作为缺陷磁通信号，其中FM＝NO×SM，SM为1个股距的采样点数量，NO为5-10的自然数；
    所述提取缺陷漏磁信号包括：根据每组漏磁采集点的位置信息，对钢丝绳的漏磁信号向前和向后提取LFM个点作为缺陷漏磁信号，LSM为1个股距的采样点数量，LFM＝LNO×LSM，LNO为5-10的自然数。
  8. 根据权利要求7所述的检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于：
    步骤S50中计算被测钢丝绳的缺陷宽度flw包括以下步骤：
    步骤S51，先根据公式h(s)＝df(s)/ds(s＝1,2,…,k)求解被测钢丝绳的缺陷磁通信号的微分结果h(s)，其中，k为缺陷磁通信号的数据个数，f(s)为缺陷磁通信号的数据；然后根据缺陷磁通信号的波形峰值点位置，向前取h(s)的最大值的位置，向后取h(s)的最小值的位置，计算最大值和最小值之间的间距作为所述缺陷磁通信号的波形宽度值Ylw；
    步骤S52，根据缺陷漏磁信号的波形峰值点位置，计算缺陷漏磁信号的最大值和最小值之间的间距作为所述缺陷磁通信号的波形宽度值Xlw；
    步骤S53，当|Ylw-Xlw|＜M，M为钢丝绳股距，选取Ylw、Xlw中的较大值为缺陷宽度flw；当|Ylw-Xlw|≥M，缺陷宽度flw＝(Ylw+Xlw)/2-LF，其中LF为传感器距离钢丝绳表面的距离。
  9. 根据权利要求8所述的检测及区分钢丝绳内外缺陷的无损检测方法，其特征在于：
    步骤S60，计算被测钢丝绳的缺陷截面损失量fs包括如下步骤：
    步骤S61，将得到的缺陷磁通信号的波形峰值T和波形基线值L，通过公式VPP＝|T–L|计算所述缺陷磁通信号的波形峰峰值VPP；
    步骤S62，设计关系函数fs＝f1(VPP,flw)，其中，f1为训练的多次方程组或者多层神经 网络；
    步骤S63，将步骤S50中得到的缺陷宽度值flw和S61的缺陷磁通信号的波形峰峰值VPP，代入到步骤S62中的多次方程组或者多层神经网络中，计算得到钢丝绳的缺陷截面损失量fs；
    其中，步骤S62中所述训练的多次方程组或者多层神经网络f1采用以下步骤得到：
    步骤S621，设计x个缺陷宽度，y个缺陷截面损失量，总共x×y个标准表面缺陷，x和y均为自然数；
    步骤S622：对x×y个标准表面缺陷经过步骤S10至步骤S60的计算得到相应的缺陷磁通波形峰峰值；
    步骤S623：将相应的缺陷磁通波形峰峰值和标准表面缺陷的缺陷宽度作为输入自变量，缺陷截面损失量作为输出标准量，训练得到多次方程组或者多层神经网络f1。
  10. 一种检测及区分钢丝绳内外缺陷的无损检测装置，其特征在于，其包括：
    励磁结构，用于将钢丝绳励磁至饱和或近似饱和状态；
    磁通检测传感器，用于获取被测钢丝绳的磁通信号；
    磁场强度检测传感器，用于获取所述被测钢丝绳的漏磁信号；
    以及信号采集和处理系统，所述信号采集和处理系统采用如权利要求2～9任意一项所述的检测及区分钢丝绳内外缺陷的无损检测方法进行处理；
    所述信号采集和处理系统包括信号采集单元、信号预处理单元、缺陷位置计算单元、缺陷信号提取单元、缺陷宽度计算单元、缺陷截面损失量计算单元、内外部缺陷区分单元和缺陷埋藏深度计算单元，其中，
    所述信号采集单元用于采集被测钢丝绳的磁通信号和漏磁信号，
    所述信号预处理单元用于对被测钢丝绳的磁通信号和漏磁信号进行预处理；
    所述缺陷位置计算单元用于根据预处理后的磁通信号、漏磁信号分别与预设磁通信号阈值、预设漏磁信号阈值进行对比，计算缺陷位置；
    所述缺陷信号提取单元用于根据缺陷位置，提取缺陷磁通信号和缺陷漏磁信号；
    所述缺陷宽度计算单元用于根据缺陷磁通信号和缺陷漏磁信号，计算被测钢丝绳的缺陷宽度flw；
    所述缺陷截面损失量计算单元用于根据被测钢丝绳的缺陷宽度flw，计算被测钢丝绳的缺陷截面损失量fs；
    所述内外部缺陷区分单元用于根据被测钢丝绳的缺陷宽度flw和缺陷磁通信号的波形峰峰值FV，区分是内部缺陷还是外部缺陷；
    所述缺陷埋藏深度计算单元用于根据被测钢丝绳的缺陷宽度flw、被测钢丝绳的缺陷截面损失量fs和虚拟截面损失量ffs计算缺陷埋藏深度。
PCT/CN2021/081320 2020-06-11 2021-03-17 一种检测及区分钢丝绳内外缺陷的无损检测方法和装置 WO2021248962A1 (zh)

Priority Applications (2)

Application Number Priority Date Filing Date Title
EP21758301.2A EP3951381B1 (en) 2020-06-11 2021-03-17 Non-destructive testing method and device for testing and distinguishing internal and external defects of steel wire rope
US17/436,086 US11703477B2 (en) 2020-06-11 2021-03-17 Nondestructive testing method and device for detecting and distinguishing internal defect and external defect of wire rope

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN202010529779.1 2020-06-11
CN202010529779.1A CN111579637B (zh) 2020-06-11 2020-06-11 一种检测及区分钢丝绳内外缺陷的无损检测方法和装置

Publications (1)

Publication Number Publication Date
WO2021248962A1 true WO2021248962A1 (zh) 2021-12-16

Family

ID=72125810

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2021/081320 WO2021248962A1 (zh) 2020-06-11 2021-03-17 一种检测及区分钢丝绳内外缺陷的无损检测方法和装置

Country Status (4)

Country Link
US (1) US11703477B2 (zh)
EP (1) EP3951381B1 (zh)
CN (1) CN111579637B (zh)
WO (1) WO2021248962A1 (zh)

Families Citing this family (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN111579637B (zh) * 2020-06-11 2022-04-29 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 一种检测及区分钢丝绳内外缺陷的无损检测方法和装置
JP7371791B2 (ja) * 2020-09-10 2023-10-31 株式会社島津製作所 ワイヤロープ検査装置およびワイヤロープ検査システム
CN112730602B (zh) * 2020-12-11 2024-04-23 浙江大学 生物安全柜前窗移门悬挂钢丝缺陷的检测装置和方法
CN114720552A (zh) * 2022-03-29 2022-07-08 哈尔滨理工大学 一种识别钢丝绳缺陷的电磁检测方法
KR102607277B1 (ko) * 2023-06-07 2023-11-29 한전케이피에스 주식회사 결함 감지 시스템 및 그 방법

Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007012331A2 (de) * 2005-07-29 2007-02-01 V & M Deutschland Gmbh Verfahren zur zerstörungsfreien prüfung von rohren auf oberflächenfehler
CN101216460A (zh) * 2008-01-11 2008-07-09 清华大学 基于交直流复合磁化的漏磁检测内外壁缺陷的识别方法
CN102798660A (zh) * 2012-08-30 2012-11-28 东北大学 基于三轴漏磁与电涡流的管道内外壁缺陷检测装置及方法
CN102798667A (zh) * 2012-08-16 2012-11-28 山东省科学院激光研究所 激光-电磁超声无损检测系统的金属缺陷检测方法
CN104514987A (zh) * 2014-12-19 2015-04-15 清华大学 管道三维漏磁成像缺陷量化方法
CN110568059A (zh) * 2019-09-24 2019-12-13 哈尔滨工业大学（深圳） 钢丝绳无损检测方法及装置
CN111024807A (zh) * 2019-12-18 2020-04-17 哈尔滨工业大学（深圳） 一种钢丝绳磁通检测损伤定量方法
CN111579637A (zh) * 2020-06-11 2020-08-25 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 一种检测及区分钢丝绳内外缺陷的无损检测方法和装置

Family Cites Families (12)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
US4659991A (en) * 1983-03-31 1987-04-21 Ndt Technologies, Inc. Method and apparatus for magnetically inspecting elongated objects for structural defects
KR950013780B1 (ko) * 1990-05-22 1995-11-16 후루까와 뗀끼 고요교오 가부시끼가이샤 이동하고 있는 길다란 물건의 길이를 측정하는 장치 및 그 방법
US5565771A (en) * 1995-01-18 1996-10-15 Noranda, Inc. Apparatus for increasing linear resolution of electromagnetic wire rope testing
US5804964A (en) * 1996-11-29 1998-09-08 Noranda Inc. Wire rope damage index monitoring device
GB2492745A (en) * 2011-06-06 2013-01-16 Silverwing Uk Ltd Magnetic flux leakage inspection
CN103703361B (zh) * 2011-08-18 2017-06-20 贝卡尔特公司 股绳中的磁通量泄漏误差探测
US9103798B2 (en) * 2011-12-07 2015-08-11 Ndt Technologies, Inc. Magnetic inspection device and method
KR101482347B1 (ko) * 2012-12-27 2015-01-13 주식회사 포스코 강판의 내부 결함 검출 장치 및 방법
DE102014212499A1 (de) * 2014-06-27 2015-12-31 Institut Dr. Foerster Gmbh & Co. Kg Verfahren und Vorrichtung zur Streuflussprüfung
RU2697061C1 (ru) * 2015-08-31 2019-08-09 ДжейАрБи ИНДЖИНИРИНГ ПТИ ЛТД Способ и система для выявления нарушения непрерывности материала в намагничиваемом изделии
CN108828059B (zh) * 2018-06-29 2020-04-07 清华大学 电磁多场耦合缺陷综合检测评价方法及装置
CN110006338B (zh) * 2019-04-28 2020-11-06 哈尔滨工业大学（深圳） 一种钢丝绳损伤面积检测方法

Patent Citations (8)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
WO2007012331A2 (de) * 2005-07-29 2007-02-01 V & M Deutschland Gmbh Verfahren zur zerstörungsfreien prüfung von rohren auf oberflächenfehler
CN101216460A (zh) * 2008-01-11 2008-07-09 清华大学 基于交直流复合磁化的漏磁检测内外壁缺陷的识别方法
CN102798667A (zh) * 2012-08-16 2012-11-28 山东省科学院激光研究所 激光-电磁超声无损检测系统的金属缺陷检测方法
CN102798660A (zh) * 2012-08-30 2012-11-28 东北大学 基于三轴漏磁与电涡流的管道内外壁缺陷检测装置及方法
CN104514987A (zh) * 2014-12-19 2015-04-15 清华大学 管道三维漏磁成像缺陷量化方法
CN110568059A (zh) * 2019-09-24 2019-12-13 哈尔滨工业大学（深圳） 钢丝绳无损检测方法及装置
CN111024807A (zh) * 2019-12-18 2020-04-17 哈尔滨工业大学（深圳） 一种钢丝绳磁通检测损伤定量方法
CN111579637A (zh) * 2020-06-11 2020-08-25 哈尔滨工业大学(深圳)(哈尔滨工业大学深圳科技创新研究院) 一种检测及区分钢丝绳内外缺陷的无损检测方法和装置

Non-Patent Citations (1)

* Cited by examiner, † Cited by third party
Title
See also references of EP3951381A4 *

Also Published As

Publication number Publication date
CN111579637A (zh) 2020-08-25
EP3951381C0 (en) 2023-06-07
CN111579637B (zh) 2022-04-29
US11703477B2 (en) 2023-07-18
US20220187246A1 (en) 2022-06-16
EP3951381A1 (en) 2022-02-09
EP3951381B1 (en) 2023-06-07
EP3951381A4 (en) 2022-05-18

Similar Documents

Publication Publication Date Title
WO2021248962A1 (zh) 一种检测及区分钢丝绳内外缺陷的无损检测方法和装置
CN109782274B (zh) 一种基于探地雷达信号时频统计特征的水损害识别方法
CN103499585B (zh) 基于机器视觉的非连续性锂电池薄膜缺陷检测方法及其装置
EP2767228B1 (en) Device and method for improving identification accuracy of fetal heart rate deceleration
CN102928435A (zh) 基于图像和超声信息融合的飞机蒙皮损伤识别方法及装置
Zhang et al. Quantitative detection of wire rope based on three-dimensional magnetic flux leakage color imaging technology
CN110568059B (zh) 钢丝绳无损检测方法及装置
CN106447040A (zh) 基于异类多传感器数据融合的机械设备健康状态评估方法
CN109387565A (zh) 一种通过分析声音信号检测刹车片内部缺陷的方法
Fu et al. Research on image-based detection and recognition technologies for cracks on rail surface
CN110032988A (zh) 紫外拉曼光谱系统实时降噪增强方法
Li et al. A new quantitative non-destructive testing approach of broken wires for steel wire rope
CN115345876A (zh) 一种基于超声图像的螺栓螺纹缺陷检测方法
CN106404893B (zh) 一种轴向漏磁缺陷信号自动检测方法
CN102988041A (zh) 心磁信号噪声抑制中的信号选择性平均方法
Pan et al. Adaptive local flaw detection based on magnetic flux leakage images with a noise distortion effect for steel wire ropes
CN114332081B (zh) 基于图像处理的纺织品表面异常判定方法
CN101718747A (zh) 一种利用金属磁记忆检测判别铁磁材料应力集中的方法
DE102005063352B4 (de) Verfahren zur zerstörungsfreien Prüfung von Rohren auf Oberflächenfehler
CN111189923A (zh) 冶金板材缺陷的变频率比较定量检测方法
Yao et al. Research on wavelet denoising method based on soft threshold in wire rope damage detection
CN109459489B (zh) 一种基于磁记忆检测原理的电梯裂纹检测方法
KR101492254B1 (ko) 정도관리를 위한 초음파 진단 장치 및 초음파 진단 방법
CN105513058A (zh) 一种脑激活区检测方法和装置
CN111024807B (zh) 一种钢丝绳磁通检测损伤定量方法

Legal Events

Date Code Title Description
ENP Entry into the national phase

Ref document number: 2021758301

Country of ref document: EP

Effective date: 20210831

NENP Non-entry into the national phase

Ref country code: DE