CN113702513B - Method for identifying metal material based on predictive function model - Google Patents
Method for identifying metal material based on predictive function model Download PDFInfo
- Publication number
- CN113702513B CN113702513B CN202110806763.5A CN202110806763A CN113702513B CN 113702513 B CN113702513 B CN 113702513B CN 202110806763 A CN202110806763 A CN 202110806763A CN 113702513 B CN113702513 B CN 113702513B
- Authority
- CN
- China
- Prior art keywords
- signal
- standard
- metal sample
- linear prediction
- sample
- Prior art date
- Legal status (The legal status 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 status listed.)
- Active
Links
- 238000000034 method Methods 0.000 title claims abstract description 50
- 239000007769 metal material Substances 0.000 title claims abstract description 33
- 239000000523 sample Substances 0.000 claims abstract description 170
- 239000002184 metal Substances 0.000 claims abstract description 120
- 229910052751 metal Inorganic materials 0.000 claims abstract description 120
- 238000005070 sampling Methods 0.000 claims description 55
- 238000005311 autocorrelation function Methods 0.000 claims description 22
- 239000000463 material Substances 0.000 claims description 15
- 238000004364 calculation method Methods 0.000 claims description 12
- 239000011159 matrix material Substances 0.000 claims description 10
- 238000009826 distribution Methods 0.000 claims description 5
- 125000003275 alpha amino acid group Chemical group 0.000 claims description 3
- 150000001875 compounds Chemical class 0.000 claims description 3
- 238000009659 non-destructive testing Methods 0.000 abstract description 3
- 238000005516 engineering process Methods 0.000 abstract description 2
- 230000000149 penetrating effect Effects 0.000 abstract description 2
- 230000035945 sensitivity Effects 0.000 abstract description 2
- 150000002739 metals Chemical class 0.000 description 7
- 238000002474 experimental method Methods 0.000 description 5
- XLYOFNOQVPJJNP-UHFFFAOYSA-N water Substances O XLYOFNOQVPJJNP-UHFFFAOYSA-N 0.000 description 5
- 239000013078 crystal Substances 0.000 description 4
- 238000001514 detection method Methods 0.000 description 4
- 230000002159 abnormal effect Effects 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 3
- 238000005520 cutting process Methods 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 229910001220 stainless steel Inorganic materials 0.000 description 3
- 239000010935 stainless steel Substances 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- 230000008878 coupling Effects 0.000 description 2
- 239000007822 coupling agent Substances 0.000 description 2
- 238000010168 coupling process Methods 0.000 description 2
- 238000005859 coupling reaction Methods 0.000 description 2
- 238000013500 data storage Methods 0.000 description 2
- 238000010586 diagram Methods 0.000 description 2
- 238000011160 research Methods 0.000 description 2
- 238000004448 titration Methods 0.000 description 2
- 230000009286 beneficial effect Effects 0.000 description 1
- 238000012512 characterization method Methods 0.000 description 1
- 238000007796 conventional method Methods 0.000 description 1
- 230000007547 defect Effects 0.000 description 1
- 230000001066 destructive effect Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004442 gravimetric analysis Methods 0.000 description 1
- 238000009434 installation Methods 0.000 description 1
- 238000011900 installation process Methods 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 239000000203 mixture Substances 0.000 description 1
- 238000010606 normalization Methods 0.000 description 1
- 238000000053 physical method Methods 0.000 description 1
- 238000007781 pre-processing Methods 0.000 description 1
- 239000010970 precious metal Substances 0.000 description 1
- 239000002994 raw material Substances 0.000 description 1
- 230000001953 sensory effect Effects 0.000 description 1
- 238000001228 spectrum Methods 0.000 description 1
- 239000000126 substance Substances 0.000 description 1
- 238000006467 substitution reaction Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N29/00—Investigating or analysing materials by the use of ultrasonic, sonic or infrasonic waves; Visualisation of the interior of objects by transmitting ultrasonic or sonic waves through the object
- G01N29/44—Processing the detected response signal, e.g. electronic circuits specially adapted therefor
- G01N29/4454—Signal recognition, e.g. specific values or portions, signal events, signatures
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/01—Indexing codes associated with the measuring variable
- G01N2291/015—Attenuation, scattering
-
- G—PHYSICS
- G01—MEASURING; TESTING
- G01N—INVESTIGATING OR ANALYSING MATERIALS BY DETERMINING THEIR CHEMICAL OR PHYSICAL PROPERTIES
- G01N2291/00—Indexing codes associated with group G01N29/00
- G01N2291/02—Indexing codes associated with the analysed material
- G01N2291/023—Solids
- G01N2291/0234—Metals, e.g. steel
Landscapes
- Engineering & Computer Science (AREA)
- Signal Processing (AREA)
- Physics & Mathematics (AREA)
- Health & Medical Sciences (AREA)
- Life Sciences & Earth Sciences (AREA)
- Chemical & Material Sciences (AREA)
- Analytical Chemistry (AREA)
- Biochemistry (AREA)
- General Health & Medical Sciences (AREA)
- General Physics & Mathematics (AREA)
- Immunology (AREA)
- Pathology (AREA)
- Investigating Or Analyzing Materials By The Use Of Ultrasonic Waves (AREA)
Abstract
The invention belongs to the technical field of metal ultrasonic nondestructive testing, and particularly relates to a method for identifying a metal material based on a prediction function model. The invention provides a new method for metal identification, which is to apply the ultrasonic nondestructive testing technology based on linear prediction coefficient algorithm to the field of metal identification for the first time; the method only needs to enable the ultrasonic wave transmitting probe to contact the detected object from one side, and has the advantages of simple and safe operation, light equipment, low cost, strong penetrating capacity, good directivity, high sensitivity and the like.
Description
Technical Field
The invention belongs to the technical field of ultrasonic nondestructive testing, relates to a method for identifying a metal material by using a linear prediction coefficient, and particularly relates to a method for identifying a metal material based on a prediction function model.
Background
In recent years, raw material prices are continuously increased, precious metal materials are relatively short, and in order to obtain violence, partial bad merchants use cheap metals to replace expensive metals, so that the production and life of people are greatly influenced. Therefore, the realization of the true and false identification has wide application in various fields such as industry, military, cultural relics identification and the like.
Currently, there are many methods for distinguishing metals, such as physical methods: sensory recognition, fracture recognition, spark recognition, and the like; the chemical method comprises the following steps: titration analysis methods, gravimetric analysis methods, volumetric analysis methods, and the like. However, the above methods are destructive to the metal itself, and some of them are complicated to operate and difficult to apply in a wide range.
The ultrasonic nondestructive identification has the characteristics of nondestructive performance, simplicity in operation, direct and rapid operation, high accuracy, wide application range and the like, and becomes an important identification mode in the detection field. When the ultrasonic wave propagates in the metal material, the influence of the metal crystal grains causes scattering, and the characteristic quantity extracted by the reflection and scattering signals carrying the information of the metal crystal grains is analyzed, so that the identification can be realized.
In the existing research technology, an ultrasonic attenuation spectrum correlation coefficient method, a weighted Euclidean distance method and the like are adopted to identify a metal sample, but the existing research has certain limitations, such as: because the probe has directivity, when gathering the signal, must strict control place the probe in same position and same direction, in actual operation, this is difficult to do, if one of them gathers wrong, and the phase difference is great with other signals, has very big influence to whole experimental result, therefore, its accuracy to the operation and the requirement of probe are higher.
Disclosure of Invention
Aiming at the technical problems in the prior art, the invention provides a method for identifying metal materials based on a predictive function model, which has the advantages of easy operation, strong penetrating power, good directionality and high sensitivity, can improve the identification accuracy of the metal materials, and reduces the dependence on people and detection probes.
In order to achieve the above purpose, the technical scheme adopted by the invention is as follows:
a method of identifying a metallic material based on a predictive function model includes the steps of:
(1) The ultrasonic probe respectively collects time domain signals of a standard metal sample and a metal sample to be identified, and sequentially marks the time domain signals as the standard signal and the signal to be identified;
(2) Taking a time domain signal of a standard metal sample, constructing a prediction function model, calculating a prediction error, and obtaining a p-order linear prediction coefficient a i of the standard metal sample by using a minimum mean square error criterion and an iterative algorithm, wherein i=1-p;
(3) According to the type of the metal sample and the p-order linear prediction coefficient a i, determining a threshold delta of the standard signal by adopting a multiple standard deviation method;
(4) Calculating the p-order linear prediction coefficient of each signal to be identified according to the method of the step (2);
(5) Comparing the p-order linear prediction coefficient of the signal to be identified with the threshold delta in the step (3), and if the p-order linear prediction coefficients of the signal to be identified are all within the threshold delta, obtaining the same material; if a certain level or a certain number of levels exceed the threshold range, the metal material is the dissimilar material, and the identification of the metal material is realized.
Further, the step (2) specifically includes:
(2.1) acquiring a time domain signal of a standard metal sample by using an ultrasonic probe, wherein the amplitude is represented as X (n), and n is an ultrasonic signal sampling point;
(2.2) normalizing the amplitude X (n) of each sampling point of the time domain signal of the standard metal sample:
Wherein: x (n) min is the minimum amplitude in the echo signal; x (n) max is the maximum amplitude in the echo signal, and X (n) is the signal amplitude of the nth sampling point of the standard metal sample signal, namely the real signal;
(2.3) obtaining the signal amplitude of the nth sampling point of the standard metal sample signal by the prediction function according to the standard metal sample signal amplitude of the p sampling points before the nth sampling point Namely, standard metal sample ultrasonic prediction signals;
(2.4) calculating a prediction error e (n) between the ultrasonic predicted signal and the true signal of the metal sample and a mean square error epsilon:
(2.5) calculating a i by using a minimum mean square error criterion, enabling epsilon to take a minimum value, and obtaining a linear equation system with a i as a variable:
Wherein a i is a linear prediction coefficient; x (n-j) represents the standard metal sample signal amplitude of j sampling points before the nth sampling point;
(2.6) constructing an autocorrelation function R n (j) of the time domain signal of the standard metal sample from the linear equation set in the step (2.5), wherein N is the total number of sampling data of the echo signal of the metal sample acquired by using the ultrasonic probe, so that N is more than or equal to 0 and less than or equal to N:
Wherein: j is the time delay of the autocorrelation function of the ultrasonic signal of the metal sample, R n (i-j) represents the jth autocorrelation function value predicted by using the i autocorrelation function values before the jth sampling point;
(2.7) splitting R n (j) into Toeplize matrices and calculating coefficients by iterative algorithm:
A i is calculated iteratively, expressed as follows:
In the formula (8), the amino acid sequence of the compound, The method is a solution of Toeplize matrix formed by an autocorrelation function matrix of the time domain signal of the metal sample and a prediction mean square error matrix of the time domain signal of the metal sample; calculating coefficients by an iterative algorithm; k j is the reflection coefficient, l represents the order of each iteration calculation of the linear prediction coefficient of the metal sample time domain signal; /(I)For the result obtained before the current iteration order, R n (i-j) represents the jth autocorrelation function value predicted using the i autocorrelation function values before the jth sampling point; epsilon (j-1) represents the predicted mean square error of the metal sample time domain signal calculated in the previous iteration of the current iteration when the iteration is calculated.
Further, the step (3) specifically includes:
The distribution in signal acquisition is simulated based on gaussian distribution, and a threshold delta is calculated according to the following formula (9), specifically:
wherein m is the number of standard signals; a r is a p-order linear prediction coefficient; recording device Representing the average value of p-order linear prediction coefficients of m standard signals; δ=standard deviation of p-order linear prediction coefficients of m standard signals.
Further, in the step 3), if the p-order linear prediction coefficient a i of the standard signal exceeds the range of the threshold value Δ according to the calculated threshold value Δ, the standard signal is described as abnormal collection, the standard signal is removed by using a program algorithm, and the threshold value Δ of the standard signal is recalculated after the standard signal is removed.
Further, the conditions for acquiring the time domain signal in the step (1) are as follows: the central frequency of the ultrasonic receiving/transmitting probe is 5-6MHz, and the diameter of the wafer is 10mm; the pulse repetition frequency of the ultrasonic pulse transmitting/receiving instrument is 100Hz, the pulse voltage is 100V, and the gain is +8dB; the sampling frequency of the digital oscilloscope is 5G S/s, the sampling time is 20 mu s, and the sampling average frequency is 2000 times.
Further, in the step (1), each time a signal is acquired, the probe needs to be lifted and replaced at the same position.
Further, the number of times of acquisition of the standard signal is not less than 20 times, and the number of times of acquisition of the signal to be identified is not less than 5 times.
Compared with the prior art, the invention has the beneficial effects that:
1. According to the invention, the time domain ultrasonic reflection and scattering signals in the metal material are obtained, the p-order linear prediction coefficient of the signals is calculated and used as the characteristic quantity of ultrasonic identification, and the threshold value is determined by multiple standard deviations, so that the method is a novel method for realizing metal identification, errors in metal detection are greatly reduced, and the accuracy of identifying metals is improved.
2. The method overcomes the defects of the traditional metal material distinguishing method, has no damage to metal, has the characteristics of simple operation, low cost, high efficiency, wide application range, accurate and reasonable recognition result and the like, and is suitable for popularization and application.
3. According to the method, the identification program software is written according to the provided algorithm, so that the operation is simple, and the data storage is convenient and quick; the display of the identification result is also increased, so that the analysis is intuitive and convenient; and meanwhile, the program software is convenient and quick to install.
Drawings
FIG. 1 is a diagram of an identification procedure operation interface;
Fig. 2 is an echo signal diagram of a metal standard sample.
Detailed Description
The invention will now be described in detail with reference to the drawings and examples.
The invention provides a method for identifying a metal material based on a linear prediction coefficient, which comprises the following steps:
(1) The ultrasonic probe collects time domain signals of a standard metal sample and a metal sample to be identified;
the connection relation of the ultrasonic anti-counterfeiting recognition device is as follows: the ultrasonic pulse transmitting/receiving instrument is connected with a computer through a digital oscilloscope, and the receiving/transmitting probe is connected with the ultrasonic pulse transmitting/receiving instrument and then placed on the surface of a metal sample to be identified, so that the receiving/transmitting probe is in coupling contact with the metal sample through a coupling agent.
The central frequency of the ultrasonic receiving/transmitting probe is 5-6MHz, and the diameter of the wafer is 10mm; the pulse repetition frequency of the ultrasonic pulse transmitting/receiving instrument is 100Hz, the pulse voltage is 100V, and the gain is +8dB; the sampling frequency of the digital oscilloscope is 5G S/s, the sampling time is 20 mu s, and the sampling average frequency is 2000 times.
Description of probe placement: every time a signal is acquired, the probe needs to be replaced, the acquisition point can be marked in advance, and the probe is ensured to be placed at the same position on the metal surface every time. The probe can rotate a small angle, the designated direction of the probe is prescribed in advance, for example, the probe line is taken as a reference, and the acquisition of each signal is required to be within the [ -5 degrees, +5 degrees ] angle interval of the designated direction.
The couplant is water, the couplant needs to be added dropwise again when the signal is taken once, the same amount of the couplant is ensured as much as possible, and one drop of water can be taken in the experiment.
In order to obtain a correct identification result, the acquisition times of the signals can be determined according to actual conditions. However, the number of standard signal acquisitions is preferably not less than 20, the number of signal acquisitions to be identified is preferably not less than 3, and the echo signal map is acquired, as shown in fig. 2.
Acquiring a time domain signal of a standard metal sample by using an ultrasonic probe, wherein the amplitude is represented as X (n), and n is an ultrasonic signal sampling point; normalizing the amplitude X (n) of each sampling point of the standard metal sample echo signal by using the formula (1):
Wherein X (n) min is the minimum amplitude in the echo signal; x (n) max is the maximum amplitude in the echo signal, and X (n) is the signal amplitude of the nth sampling point of the standard metal sample signal, namely the real signal;
(2) According to the standard metal sample signal amplitude of p sampling points before the nth sampling point, obtaining the signal amplitude of the nth sampling point of the standard metal sample signal through a prediction function Namely, standard metal sample ultrasonic prediction signals;
Specifically, the standard metal sample signal amplitude according to p sampling points before the nth sampling point is: x (n-1), x (n-2). X (n-p), predicting the metal sample signal amplitude of the nth sampling point by a prediction function, and constructing a p-order prediction function by taking a characteristic parameter a i of the prediction function as a characteristic quantity of the metal sample ultrasonic signal
Calculating a prediction error e (n) between the ultrasonic prediction signal and the real signal of the standard metal sample,
The coefficient a i is determined by the minimum mean square error criterion using the calculated mean square error:
calculating formula (4) to derive a i and zero the derivative to obtain a i linear equation set
A i is a linear prediction coefficient; x (n-j) represents the standard metal sample signal amplitude of j sampling points before the nth sampling point;
Finally, constructing an autocorrelation function R n (j) of the signal amplitude of the metal sample by the formula (5), wherein N is the total sampling data of the echo signal of the metal sample acquired by using the ultrasonic probe, so that N is more than or equal to 0 and less than or equal to N:
a i is a linear prediction coefficient, and is a parameter which enables a prediction function to meet minimum mean square error, namely a characteristic quantity of a metal sample ultrasonic signal; wherein: j is the time delay of the autocorrelation function of the ultrasonic signal of the metal sample, R n (i-j) represents the jth autocorrelation function value predicted by using the i autocorrelation function values before the jth sampling point;
And iteratively calculating (6) the p-order linear prediction coefficient a i (i=1, 2,3,.. P.) of the metal sample time domain signal, specifically splitting R n (j) into Toeplize matrices and calculating the coefficients by an iterative algorithm:
in this embodiment, the iterative calculation is started with p=1, and at this time, the autocorrelation matrix of the time domain signal of the metal sample is,
Solving outAnd ε (1); and substituting p=2 to obtain/>And ε (2); the final iteration result is expressed as follows:
In the formula (8), the amino acid sequence of the compound, The method is a solution of Toeplize matrix formed by an autocorrelation function matrix of the time domain signal of the metal sample and a prediction mean square error matrix of the time domain signal of the metal sample; calculating coefficients by an iterative algorithm; k j is the reflection coefficient, l represents the order of each iteration calculation of the linear prediction coefficient of the metal sample time domain signal; /(I)For the result obtained before the current iteration order, R n (i-j) represents the jth autocorrelation function value predicted using the i autocorrelation function values before the jth sampling point; epsilon (j-1) represents the predicted mean square error of the metal sample time domain signal calculated in the previous iteration of the current iteration when the iteration is calculated.
The p-order linear prediction coefficient a i is taken as the characteristic quantity of the ultrasonic signal of each metal sample, k j is a reflection coefficient, l represents the order of each iterative calculation of the linear prediction coefficient of the time domain signal of the metal sample,And (3) for the result obtained before the current iteration order, starting from the first order, calculating until the total order p is the iteration.
For example, first-order iteration results inSubstituting the second order iteration and obtaining/>And/>By such a push, the p-order linear prediction coefficient of the time domain signal of the characterization metal sample is finally obtained: a i: /(I)
(3) According to the type of the metal sample and the p-order linear prediction coefficient a i, determining a threshold delta of the standard signal by adopting a multiple standard deviation method;
Specifically, according to the central limit theorem, the signal acquired by each metal sample for many times obeys Gaussian distribution, and the threshold delta of the standard signal is calculated by the Laida rule, specifically:
Wherein m is the number of standard signals; a r is the p-order linear prediction coefficient of the ultrasonic signals of m groups of repeatedly acquired metal samples; recording device The average value of the p-order linear prediction coefficients of m groups of standard signals is obtained; /(I)Is the standard deviation of the p-order linear prediction coefficients of the m groups of signals.
Meanwhile, if the p-order linear prediction coefficient a i of the standard signal exceeds the range of the threshold delta according to the calculated threshold delta, the standard signal is abnormal in acquisition, the standard signal is removed by using a program algorithm, and the threshold delta of the standard signal is recalculated after the standard signal is removed.
In this embodiment, the multiple standard deviation is calculated by adding and subtracting several times of the standard deviation from the average value. The selection of the standard deviation multiple is related to the metal material, if the scattering of the internal crystal grains of the standard metal sample on the ultrasonic wave is weaker, the 3 times is more accurate, otherwise, if the scattering of the internal crystal grains of the metal material on the ultrasonic wave is stronger, the 5 times is more capable of improving the fault tolerance and reducing the false recognition risk caused by the operation error of personnel. Several hundred groups of experiments are carried out on various metals, and the universality is the best when the number is 3-5 times, so that the recognition of most metals can be satisfied.
In this embodiment, the multiple of the standard deviation is 3, and the standard deviation just meets the raydad criterion. However, in practical application, certain fault tolerance and universality of different metal materials are required in consideration of unavoidable operation errors, so that improvement is performed on the basis of the Laida criterion, namely the multiple can be manually changed according to the self-characteristics of the metal materials, and the accuracy and the fault tolerance are further adjusted.
In the present embodiment of the present invention,
(4) Calculating the p-order linear prediction coefficient of each signal to be identified according to the method of the step (2);
(5) Comparing the p-order linear prediction coefficient of the signal to be identified with the threshold delta in the step (3), and if the p-order linear prediction coefficient of the signal to be identified is within the threshold delta, identifying the result as the same sample; if the coefficients of 1 order and above in the p-order linear prediction coefficients of the signal to be detected are out of the range of the threshold delta, the identification result is different samples. A method of identifying a metal material based on a linear prediction coefficient is realized.
Referring to fig. 1, the present invention is written as an identification program according to the above mentioned algorithm, and the operation can complete the identification only in 3 steps: step 1, clicking a button of a standard metal signal, and selecting a folder for storing the standard metal sample signal; step 2, clicking a button of 'metal signal to be identified', and selecting a folder for storing the metal sample signal to be identified; and 3, clicking an identification button to obtain a result. If the signal to be identified comes from the standard metal sample, a green result frame shown in fig. 1 is obtained, and the same metal object is displayed; if the signal to be identified is not from a standard metal sample, the result box will turn red and display a "different metal item".
The recognition program provided by the invention further increases the specific display of the recognition result, and each time the recognition button shown in fig. 1 is clicked, the final recognition result, the number of the p-order linear prediction coefficients meeting the threshold value and the number of the p-order linear prediction coefficients exceeding the threshold value are given out for further analysis.
Because the prior art has complicated requirements on the format of the signal storage folder, 20 empty folders are required to be established firstly for each storage, and then the acquired time domain signals of the 20 standard metal samples in the csv format are sequentially put into the 20 empty folders. The recognition program software provided by the invention optimizes the signal storage mode, can automatically recognize the number of data stored by the probe put down each time and automatically perform algorithm processing, is convenient and quick in data storage, only needs to distinguish sample folders, and does not need to perform complicated folder processing.
Meanwhile, the identification program provided by the invention uses the compiler to compile the executable file, can be directly used on any Windows7 or above operating system computer, has the executable file size not exceeding 1GB, is convenient and quick in installation process, and can be installed only by selecting an installation catalog.
Example 1: identification between dissimilar metallic materials of similar composition
Experimental samples: the three stainless steel round metal samples are identical in specification, similar in components and 15mm in thickness, and are respectively Cr17Ni2 (sample No. 1), 2Cr13 (sample No. 2) and 3Cr13 (sample No. 3). During the experiment, a sample No. 1 is selected as a standard sample, samples No. 1, no. 2 and No. 3 are taken as samples to be identified, and a right-angle locating plate is attached to the surface of the sample before the experiment, so that the same placement position of a probe during each signal acquisition is ensured.
In order to ensure the same experimental conditions as far as possible, the couplant (water) needs to be added dropwise again every time a signal is taken, and the dosage is one drop.
Instrument device connection: the ultrasonic pulse transmitting/receiving instrument of Panametrics-NDT 5077PR is connected with a computer through a digital oscilloscope of Tektronix-DPO5034B, and a receiving/transmitting probe with the center frequency of 5MHz is connected with the ultrasonic pulse transmitting/receiving instrument and then placed on the surface of a metal sample to be identified, so that the receiving/transmitting probe is in coupling contact with the metal sample through a coupling agent (water). Referring to fig. 2, an echo signal is received after a probe transmits an ultrasonic pulse signal, and the echo signal is sampled by an oscilloscope connected to an ultrasonic pulse transmitter/receiver.
The method for identifying a metal material based on a linear prediction coefficient provided by the embodiment comprises the following steps:
step one: collecting time domain signals of a standard metal sample and a metal sample to be identified according to a conventional method
Firstly, a couplant (two drops of water) is dripped at the right angle of the right angle locating piece, the probe is placed at the right angle inflection point of the right angle locating piece, the probe line is aligned with the right angle inflection point, and the position and the angle of each probe are ensured to be consistent. Applying a certain pressure to the probe to ensure that the probe is tightly attached to the surface of the sample and the pressure applied to the probe is the same each time so as to acquire a stable echo signal; every time a signal is taken, the probe needs to be replaced and the couplant is dripped again, and the dosage of the couplant is kept as consistent as possible each time.
Standard signal acquisition: to obtain more information of the metal grains at this point, 20 standard signals are acquired at the acquisition point. The standard signal is collected in the specified direction [ -5 DEG, +5 DEG ] angle range by taking the human operation error into consideration, so that the metal grain characteristics of the standard signal can be reflected, the abnormal signal characteristics can not be brought, and a reasonable standard signal sample database is constructed.
Collecting signals to be identified: and the method is the same as the standard signal acquisition mode and experimental conditions, and 5 times of signals to be identified are acquired within the range of the angles of minus 5 degrees and plus 5 degrees in the specified direction.
Step two: calculating the p-order linear prediction coefficient of standard signal
① The method comprises the steps of cutting off the initial wave of 20 standard signal data samples, wherein the amplitude of the initial wave signal is larger, and the initial wave of each signal is the same, so that the initial wave is cut off after data is imported to improve the signal diversity;
② Collecting a time domain signal of a standard sample by using an ultrasonic probe, wherein the amplitude is represented as X (n), and n is an ultrasonic signal sampling point; normalizing the amplitude X (n) of each sampling point of the echo signal of the standard sample by using the formula (1):
Wherein X (n) min is the minimum amplitude in the echo signal; x (n) max is the maximum amplitude in the echo signal; x (n) is the amplitude of the sampling point of the normalization processing, and is also the signal amplitude of the nth sampling point of the standard metal sample signal, namely the real signal;
In this embodiment, since the sampling point is 488, there are 438 echo amplitudes X (n) of the metal samples after cutting 50 initial wave data points, and the data amount is large, the signal amplitudes of the standard metal sample signals of 20 sampling points are obtained after normalizing the amplitudes of the first 20 sampling points as samples ：213.3333;177.3333;105.3333;65.6667;40.0000;34.0000;35.6667;53.0000 73.6667;98.0000;100.6667;96.0000;66.0000;36.3333;9.0000;20.0000 27.0000;27.3333;23.3333;11.3333, that is, the real signal is ：0.8577;0.7128;0.4228;0.2631;0.1597;0.1356;0.1423;0.2121;0.2953;0.3933;0.4040;0.3852;0.2644;0.1450;0.0349;0.0792;0.1074;0.1087;0.0926;0.0443
③ Calculating p-order linear prediction coefficients and prediction mean square errors of 20 groups of standard sample signals;
When selecting, the selection of p is determined by two points: firstly, the overall mean square error between a predicted signal and a real signal is required to be as small as possible; secondly, the calculation efficiency and the fault tolerance are required to be high. In the embodiment, when a large number of experiments are conducted to summarize that p is taken to be 12, the universal effect on various metals is good, and the mean square error of most of predicted signals is only 1% of that of real signals. The calculation speed is high, and the calculation can be completed in about 1 second. Therefore, p in this embodiment takes the 12 th order;
Since the sampling point is 488, there are 438 echo amplitudes X (n) of the metal sample after cutting 50 initial wave data points, each amplitude corresponds to a predicted mean square error, and the data size is larger, so the predicted mean square error of the first 20 points is given here as an example ：0.0090;0.0904;0.0106;0.0015;0.0012;0.0107;0.0033;0.0183;0.00003;0.0056;0.0089;0.0019;0.0004;0.0003;0.0022;0.0037;0.0007;0.0005;0.0005;0.0004
④ According to the formulas (2) to (8), obtaining 12-order linear prediction coefficients of the 20 groups of standard sample signals;
In this embodiment, the average p-order linear prediction coefficients of the 20 signals acquired by the sample No.1 are respectively: -0.9051
0.2801-0.1706-0.0612-0.0045 0.1366-0.1793-0.1217 0.0866 0.0110 0.0054-0.0300
Step three: the threshold delta of the standard signal is calculated.
The order p in this embodiment is 12, so 12 thresholds are calculated, and the upper threshold limit 12 and the lower threshold limit 12 are calculated by combining the formula (9),
Sample number 1 upper threshold limit: -0.8520 0.3552-0.1033-0.0036 0.0202 0.1777
-0.1165-0.0353 0.1833 0.0837 0.0326-0.0150；
Sample No. 1 lower threshold: -0.9581 0.2050-0.2379-0.1189-0.0293 0.0954
-0.2420-0.2082-0.0100-0.0617-0.0217-0.0450；
Step four: and calculating linear prediction coefficients of the signals to be identified.
① Importing the collected 5 signals to be identified and preprocessing the data;
② Normalizing the 5 signals to be identified according to the formula (1);
③ Respectively calculating 12-order linear prediction coefficients of 5 signals to be identified according to the formulas (2) to (8);
④ Calculating the average value of the 5 groups of linear prediction coefficients; since each sampling point is sampled for a plurality of times, the 12-order linear prediction coefficient calculated each time is averaged;
1. the average value of the linear prediction coefficients of the samples to be identified 2 and 3 are shown in table 1:
Table 11, sample No. 2, sample No. 3 linear prediction coefficient mean
Step five: identification of
And respectively comparing the 12 linear prediction coefficients of the 3 samples with the upper and lower thresholds, if the 12 linear prediction coefficients are in the respective threshold ranges, identifying the to-be-identified number as the same material, otherwise identifying the to-be-identified number as the different material.
It can be seen that the average value of 12 linear prediction coefficients calculated by the 5 times of signals to be identified of sample 1 is within the threshold range of the standard signal, so that the identification result is the same material; sample No. 2 has a 7 th order linear prediction coefficient exceeding a threshold; sample No. 3 has a 9 th order linear prediction coefficient exceeding a threshold, and samples No. 2 and No. 3 are identified as heterogeneous materials, and the identification results are shown in tables 2 to 4 below.
Table 21 sample identification results
Table 32 sample identification results
Table 43 sample identification results
As can be seen from the comparison of tables 2 to 4, the method provided by the embodiment can realize the accurate identification of different metal materials with similar components and extremely small difference of metal grains.
Example 2: identification between the same metallic materials
The present example was identical to the calculation method, the identification method, etc. employed in example 1, except that the samples were three cylindrical metal samples of 2Cr13, 50mm in diameter and 15mm in thickness, respectively, and were No. 4, no. 5, and No. 6.
And (3) selecting a sample No. 4 as a standard signal, and identifying the samples No. 4, no. 5 and No. 6 to be identified. Likewise, the standard samples were collected 20 times, and each sample to be identified was collected 5 times. The acquisition mode and the identification method are the same as those of the embodiment 1.
The calculated average linear prediction coefficients of the 20 th order standard signal of sample No. 4 are respectively: -0.9830,0.3205, -0.0948, -0.1125, -0.0576,0.2070, -0.1521, -0.2180,0.1419, -0.0367,0.0423, -0.0055;
Similarly, calculating an upper threshold limit and a lower threshold limit;
Upper threshold limit: -0.9329,0.4691,0.1353,0.1046,0.0616,0.2714, -0.0319, -0.0622,0.2402,0.0142,0.0564,0.0127;
Threshold lower limit: -1.0331,0.1719, -0.3249, -0.3295, -0.1768,0.1425, -0.2723, -0.3738,0.0436, -0.0876,0.0282, -0.0237.
The average linear prediction coefficient of each signal to be identified calculated according to the same method as the embodiment is shown in table 5.
Tables 54, 5, 6 sample linear prediction coefficients
Comparing the average linear prediction coefficient of each signal with a threshold value to obtain a recognition result: the average value of 12 linear prediction coefficients calculated by the 5 times of signals to be identified of the sample No. 4 is within the threshold range of the standard signal, so that the identification result is the same material; sample No. 5 has an 8 th order linear prediction coefficient exceeding a threshold; sample No. 6 has a 6 th order linear prediction coefficient exceeding a threshold, and samples No. 5 and No. 6 are both identified as heterogeneous materials, with specific results being shown in tables 6-8.
Table 64 sample identification results
Table 75 sample identification results
Table 8 6 sample identification results
As can be seen from the comparison of the data in tables 6 to 8, the method provided by the embodiment can realize the correct identification between different articles made of the same metal material in the same batch.
Example 3: identification of same batch of metal containers
The present example was identical to the calculation method, identification method, etc. employed in example 1, except that the sample was three stainless steel metal containers processed from the same batch of the same material: the size is as follows: the diameter is 190mm, the height is 255mm, the bottom thickness is 10mm, and the samples are respectively marked as No. 7, no. 8 and No. 9.
And (3) selecting sample No. 7 as a standard signal, and identifying samples No. 7, no. 8 and No. 9 to be identified. Likewise, the standard samples were collected 20 times, and each sample to be identified was collected 5 times. The acquisition mode and the identification method are the same as those of the embodiment 1.
The calculated average linear prediction coefficients of the 20 th order standard signal of sample No. 7 are respectively: -0.9159,0.3580, -0.1352, -0.2249,0.0983,0.0517, -0.0825,0.0635, -0.2322,0.0943, -0.0132, -0.0345;
Similarly, an upper threshold limit and a lower threshold limit are calculated.
Upper threshold limit: -0.9060,0.3745, -0.1190, -0.2128,0.1175,0.0946, -0.0303,0.1164, -0.2111,0.1372,0.0139, -0.0244;
threshold lower limit: -0.9258,0.3414, -0.1513, -0.2370,0.0792,0.0087, -0.1347,0.0106, -0.2534,0.0515, -0.0404, -0.0447.
The calculated average linear prediction coefficient for each signal to be identified is shown in table 9.
Tables 9 7, 8, and 9 sample linear prediction coefficients
Comparing the amplitude root mean square ratio of each signal with a threshold value to obtain a recognition result: the average value of 12 linear prediction coefficients calculated by the 5 times of signals to be identified of the sample No. 7 is within the threshold range of the standard signal, so that the identification result is the same material; sample 8 has 12 th order linear prediction coefficients exceeding a threshold; sample No. 9 has a linear prediction coefficient of 11 th order exceeding the threshold, and samples No. 8 and No. 9 are identified as heterogeneous materials, and the specific results are shown in tables 10 to 12.
Table 10 sample No. 7 identification results
Table 11 sample No. 8 identification results
Table 12 sample No. 9 identification results
The comparison of the data in the embodiment 3 shows that the invention can realize the correct identification of the stainless steel metal containers processed in the same batch.
The method for identifying the metal material based on the linear prediction coefficient is characterized in that the linear prediction coefficient is used as a characteristic quantity to be applied to the field of anti-counterfeiting identification of the metal material for the first time, and the method can be applied to anti-counterfeiting identification of metal cultural relics and the like; the invention optimizes the aspects of signal acquisition mode, threshold calculation and the like, can realize the correct identification between different metal materials with similar components and different samples produced by the same material, calculates the linear prediction coefficient of the signal as the characteristic quantity of ultrasonic identification by acquiring the time domain ultrasonic reflection and scattering signals in the metal materials, defines the threshold value by multiple standard deviations, is a new method for realizing metal identification, greatly reduces errors in metal detection and improves the accuracy of metal identification.
The foregoing is a further detailed description of the invention in connection with the preferred embodiments, and it is not intended that the invention be limited to the specific embodiments described. It will be apparent to those skilled in the art that several simple deductions or substitutions may be made without departing from the spirit of the invention, and these should be considered to be within the scope of the invention.
Claims (4)
1. A method for identifying a metallic material based on a predictive function model, comprising the steps of:
(1) The ultrasonic probe respectively collects time domain signals of a standard metal sample and a metal sample to be identified, and sequentially marks the time domain signals as the standard signal and the signal to be identified; collecting standard signals and signals to be identified in the specified direction [ -5 DEG, +5 DEG ] angle range;
(2) Taking a standard signal, constructing a prediction function model, calculating a prediction error, and obtaining a p-order linear prediction coefficient a i of a standard metal sample by using a minimum mean square error criterion and an iterative algorithm, wherein i=1-p;
(3) According to the type of the metal sample and the p-order linear prediction coefficient a i, determining a threshold delta of the standard signal by adopting a multiple standard deviation method;
(4) Calculating the p-order linear prediction coefficient of each signal to be identified according to the method of the step (2);
(5) Comparing the p-order linear prediction coefficient of the signal to be identified with the threshold delta in the step (3), and if the p-order linear prediction coefficients of the signal to be identified are all within the threshold delta, obtaining the same material; if one or more orders exceed the threshold range, the metal material is identified as a dissimilar material;
the step (2) specifically comprises:
(2.1) acquiring a time domain signal of a standard metal sample by using an ultrasonic probe, wherein the amplitude is represented as X (n), and n is an ultrasonic signal sampling point;
(2.2) normalizing the amplitude X (n) of each sampling point of the time domain signal of the standard metal sample:
Wherein: x (n) min is the minimum amplitude in the echo signal; x (n) max is the maximum amplitude in the echo signal, and X (n) is the signal amplitude of the nth sampling point of the standard metal sample signal, namely the real signal;
(2.3) obtaining the signal amplitude of the n-th sampling point of the standard metal sample signal by the prediction function according to the standard metal sample signal amplitude x (n-i) of the p-th sampling point before the n-th sampling point Namely, standard metal sample ultrasonic prediction signals; wherein: i is any one sampling point of P sampling points;
(2.4) calculating a prediction error e (n) between the ultrasonic predicted signal and the true signal of the metal sample and a mean square error epsilon:
(2.5) calculating a i by using a minimum mean square error criterion, enabling epsilon to take a minimum value, and obtaining a linear equation system with a i as a variable:
Wherein a i is a linear prediction coefficient; x (n-j) represents the standard metal sample signal amplitude of j sampling points before the nth sampling point;
(2.6) constructing an autocorrelation function R n (j) of the time domain signal of the standard metal sample from the linear equation set in the step (2.5), wherein N is the total number of sampling data of the echo signal of the metal sample acquired by using the ultrasonic probe, so that N is more than or equal to 0 and less than or equal to N:
Wherein: j is the time delay of the autocorrelation function of the ultrasonic signal of the metal sample, R n (i-j) represents the jth autocorrelation function value predicted by using the i autocorrelation function values before the jth sampling point;
(2.7) splitting R n (j) into Toeplize matrices and computing coefficients by iterative algorithm:
A i is calculated iteratively, expressed as follows:
In the formula (8), the amino acid sequence of the compound, The method is a solution of Toeplize matrix formed by an autocorrelation function matrix of the time domain signal of the metal sample and a prediction mean square error matrix of the time domain signal of the metal sample; calculating coefficients by an iterative algorithm; k j is the reflection coefficient, l represents the order of each iteration calculation of the linear prediction coefficient of the metal sample time domain signal; /(I)For the result obtained before the current iteration order, R n (j-l) predicts the jth autocorrelation function value by using the l autocorrelation function values before the autocorrelation function value of the jth metal sample time domain signal; epsilon (j-1) represents the predicted mean square error of the time domain signal of the metal sample calculated in the previous iteration of the current iteration during iterative calculation;
the step (3) specifically comprises the following steps:
The distribution in signal acquisition is simulated based on gaussian distribution, and a threshold delta is calculated according to the following formula (9), specifically:
wherein m is the number of standard signals; a r is a p-order linear prediction coefficient; recording device Representing the average value of p-order linear prediction coefficients of m standard signals; delta is the standard deviation of p-order linear prediction coefficients of m standard signals;
In the step (3), if the p-order linear prediction coefficient a i of the standard signal exceeds the range of 3-5 times of the threshold delta according to the calculated threshold delta, the standard signal is eliminated by using a program algorithm, and the threshold delta of the standard signal is recalculated after the standard signal is eliminated.
2. The method for identifying metallic materials based on a predictive function model as set forth in claim 1, wherein the condition for acquiring the time domain signal in the step (1) is: the central frequency of the ultrasonic receiving/transmitting probe is 5-6MHz, and the diameter of the wafer is 10mm; the pulse repetition frequency of the ultrasonic pulse transmitting/receiving instrument is 100Hz, the pulse voltage is 100V, and the gain is +8dB; the sampling frequency of the digital oscilloscope is 5G S/s, the sampling time is 20 mu s, and the sampling average frequency is 2000 times.
3. The method for identifying metallic materials based on a predictive function model as recited in claim 2, wherein in said step (1), each time a signal is acquired, the probe is lifted and repositioned at the same location.
4. The method for identifying a metal material based on a predictive function model according to claim 3, wherein the number of collection times of the standard signal is not less than 20, and the number of collection times of the signal to be identified is not less than 5.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110806763.5A CN113702513B (en) | 2021-07-16 | 2021-07-16 | Method for identifying metal material based on predictive function model |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN202110806763.5A CN113702513B (en) | 2021-07-16 | 2021-07-16 | Method for identifying metal material based on predictive function model |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN113702513A CN113702513A (en) | 2021-11-26 |
| CN113702513B true CN113702513B (en) | 2024-06-25 |
Family
ID=78648764
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN202110806763.5A Active CN113702513B (en) | 2021-07-16 | 2021-07-16 | Method for identifying metal material based on predictive function model |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN113702513B (en) |
Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112162041A (en) * | 2020-09-30 | 2021-01-01 | 陕西师范大学 | Method for identifying metal material based on Gaussian distribution of amplitude root mean square value |
Family Cites Families (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| WO2002031815A1 (en) * | 2000-10-13 | 2002-04-18 | Science Applications International Corporation | System and method for linear prediction |
| JP2005338063A (en) * | 2004-04-28 | 2005-12-08 | Japan Science & Technology Agency | Apparatus for measuring physical characteristics of sample |
| CN102590338B (en) * | 2012-01-19 | 2014-04-16 | 清华大学 | Resonance peak-based ultrasonic cavitation state identification method |
| EP2927674A1 (en) * | 2014-03-31 | 2015-10-07 | Nederlandse Organisatie voor toegepast- natuurwetenschappelijk onderzoek TNO | Method and apparatus for characterizing a medium using ultrasound measurements |
| CN106992011A (en) * | 2017-01-25 | 2017-07-28 | 杭州电子科技大学 | Engineering machinery sound identification method based on MF PLPCC features |
| CN107703210A (en) * | 2017-08-31 | 2018-02-16 | 北京金风慧能技术有限公司 | workpiece identification method and device |
| CN110797033A (en) * | 2019-09-19 | 2020-02-14 | 平安科技（深圳）有限公司 | Artificial intelligence-based voice recognition method and related equipment thereof |
-
2021
- 2021-07-16 CN CN202110806763.5A patent/CN113702513B/en active Active
Patent Citations (1)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN112162041A (en) * | 2020-09-30 | 2021-01-01 | 陕西师范大学 | Method for identifying metal material based on Gaussian distribution of amplitude root mean square value |
Non-Patent Citations (1)
| Title |
|---|
| 《基于 ARM 的声纹识别模块的软件设计与实现》;高猛;中国优秀硕士学位论文全文数据库;第17-19页 * |
Also Published As
| Publication number | Publication date |
|---|---|
| CN113702513A (en) | 2021-11-26 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN112162041B (en) | Method for recognizing metal material based on Gaussian distribution of amplitude root mean square value | |
| CN101995389B (en) | Method for fast recognition of crude oil variety through near infrared spectrum | |
| CN109649432B (en) | System and method for monitoring integrity of steel rail of cloud platform based on guided wave technology | |
| CN104359389A (en) | Pulsed eddy current measuring method for measuring relative wall thickness variable of ferromagnetic component | |
| CN102252971A (en) | Rapid detection method for mango hardness | |
| Caracciolo et al. | Precipitation classification at mid-latitudes in terms of drop size distribution parameters | |
| CN110161119B (en) | Wind power blade defect identification method | |
| CN101339150B (en) | Method for determining octane number based on dielectric spectra technology | |
| Kraljevski et al. | Machine learning for anomaly assessment in sensor networks for NDT in aerospace | |
| CN102841069B (en) | Method for rapidly identifying types of crude oil by using mid-infrared spectrum | |
| CN110542723A (en) | guided wave signal sparse decomposition and damage positioning-based two-stage damage position identification method | |
| Moctezuma et al. | Performance analysis of acoustic emission hit detection methods using time features | |
| CN114330435A (en) | Composite material defect detection method based on dynamic weight wavelet coefficient deep residual network | |
| CN113702513B (en) | Method for identifying metal material based on predictive function model | |
| CN108181377B (en) | Intelligent interpretation system and method for broken PCCP steel wire | |
| Lei et al. | Ultrasonic pig for submarine oil pipeline corrosion inspection | |
| CN110308205B (en) | Metal classification method based on EMD generalized phase permutation entropy algorithm | |
| CN108195934B (en) | Ultrasonic guided wave detection frequency optimization method based on time-frequency analysis | |
| Cai et al. | Identification method of circumferential declination based on amplitude reduction of pipeline ultrasonic internal inspection signals | |
| CN102954945B (en) | A kind of method by infrared spectrum measurement acid value for crude oil | |
| CN115684363A (en) | Concrete performance degradation evaluation method based on acoustic emission signal processing | |
| CN107515251A (en) | A kind of detecting system for metal material | |
| CN107942304A (en) | Sea clutter optimal soft survey instrument and method based on drosophila optimization algorithm optimization least square method supporting vector machine | |
| CN113624710A (en) | Terahertz information identification method for liquid substance component change and application thereof | |
| CN103412048A (en) | Method for identifying multiple metals by utilizing ultrasonic non-destructive detection method |
Legal Events
| Date | Code | Title | Description |
|---|---|---|---|
| PB01 | Publication | ||
| PB01 | Publication | ||
| SE01 | Entry into force of request for substantive examination | ||
| SE01 | Entry into force of request for substantive examination | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |