CN110687208A - A datum-free Lamb wave damage monitoring method based on hyperbolic positioning - Google Patents

Abstract

The invention provides a reference-free Lamb wave damage monitoring method based on hyperbolic positioning, comprising: arranging a diamond-shaped piezoelectric excitation/sensing array on a structure to be measured; selecting a piezoelectric sensor on a diagonal line as an excitation source to excite For the excitation signal of a single Lamb wave mode, select a pair of piezoelectric sensors on another diagonal to collect the structural response signal; repeat the above steps for the remaining piezoelectric sensors on the diagonal to obtain the damage scattering signal corresponding to the structural response signal The time difference is determined, and the positive and negative of the time difference are determined; the hyperbolic positioning method is used to calculate the damage location and approximate range. The Lamb wave damage monitoring method of the present invention has a simple array form and a small number of arrays, and the experimental measurement operation is simple. By using the principle of hyperbolic damage location, the detection and monitoring of structural damage can be realized without the need for a reference signal. and the influence of changes in external conditions, which improves the accuracy of structural damage monitoring.

Description

A datum-free Lamb wave damage monitoring method based on hyperbolic positioning

technical field

The invention belongs to the field of structural health monitoring, and in particular relates to a reference-free Lamb wave damage monitoring method based on hyperbolic positioning, which is used for damage localization imaging research.

Background technique

With the continuous improvement of structural safety and reliability requirements, the online monitoring and diagnosis of structural damage has attracted more and more attention. In order to prevent the disaster or loss caused by structural damage, effective damage monitoring must be carried out on the structure. Plate structure is one of the main forms of major engineering structures, and its real-time online structural health monitoring has received great attention. Lamb wave is an elastic wave propagating in a solid plate-like structure under free boundary conditions. Because of its small attenuation, long propagation distance, and very sensitive to small damage in the structure, it has great advantages for large-area non-destructive testing of plate-like structures. Therefore, the damage monitoring technology based on active Lamb wave is the current research hotspot of plate structure health monitoring.

Among them, most of the active Lamb wave damage monitoring methods are based on the reference signal, that is, the response signal in the structural health state is used as the reference signal, and the response signal in the current state is subtracted from the reference signal to obtain the structural damage status.

However, because the acquisition time of the reference signal and the current response signal is different, the external conditions during acquisition, such as ambient temperature, structural boundary and stress conditions, and external vibration, usually change, and internal conditions, such as the performance of the sensor itself, will also be affected by temperature. Influenced by other factors, the damage scattering signal is easily submerged in the signal changes and noises caused by changes in the internal and external conditions of the structure, which not only makes it difficult to obtain accurate damage detection results, but also affects the real-time performance of online monitoring. In addition, when the reference signal is collected, if there is damage in the structure, the monitoring method based on the reference signal will not be able to extract the damage scattering signal generated by the damage, and thus cannot obtain the correct damage monitoring result.

So far, in order to overcome the shortcomings of the existing active Lamb wave damage detection methods based on reference signals, many scholars have used various methods to study the non-reference active Lamb wave damage detection technology. Among them, the existing Lamb wave damage detection technology such as time reversal [Wang Qiang, Yan Xiajun, Chen Xiaohui, Bai Li, "Lamb wave time-reversal damage probability imaging monitoring method without reference", "Journal of Instrumentation", 2013, 34 (07), pp. 149-155], Bayesian Theory [Yin Tao, Miao Ao, Wang Xiangyu, "A method for locating damage of active Lamb wave without reference signal based on Bayesian theory", Chinese Journal of Vibration Engineering, 2017, 30 (01), pp. 33-40], etc., have realized the Lamb wave damage location monitoring without reference.

In 2016, Li Dongsheng et al. disclosed a probability-based Lamb wave no-baseline damage identification method [Li Dongsheng, Jin Mengdao, "Probability-based Lamb wave no-baseline damage identification method", "Fujian Construction Technology", 2016-06, No. Pages 29-31+34], the disclosed method adopts a pulse-echo mode distributed active sensor network to collect signals, and extracts the time characteristics of damage scattering signals through time window function and continuous wavelet transform, and finally uses The hyperbolic probabilistic imaging algorithm performs damage localization and imaging, thereby realizing Lamb wave-free baseline damage identification. This method also uses the hyperbolic probability imaging algorithm to perform datum-free structural damage localization imaging. However, the existing non-reference active Lamb wave damage monitoring methods often require more piezoelectric elements to form a dual-element piezoelectric sensing array. Eliminate the direct sensing signal, highlight the damage scattered signal, and choose two receiving sensors to form a receiving probe pair for calculating the hyperbolic trajectory, so the number of sensors required is large, and the signal processing and calculation are very complex, which affects the performance of the monitoring system. Higher requirements.

Therefore, it is necessary to find a reference-free Lamb wave damage monitoring method with a simple array arrangement, fewer numbers, and simpler signal processing.

SUMMARY OF THE INVENTION

The purpose of the present invention is to provide a reference-free Lamb wave damage monitoring method based on hyperbolic positioning, so as to realize the reference-free active Lamb wave damage positioning and monitoring.

In order to solve the above problems, the present invention provides a reference-free Lamb wave damage monitoring method based on hyperbolic positioning, including:

S1: On a structure to be tested, according to the size of the required monitoring area, a diamond-shaped piezoelectric excitation/sensing array is arranged;

S2: Select a piezoelectric sensor located on a diagonal line of the rhombus piezoelectric excitation/sensing array as the excitation source, excite the excitation signal of a single Lamb wave mode in the structure to be tested, and select the pair with the A pair of piezoelectric sensors on a straight line perpendicular to the angle are used as receivers, respectively collecting a pair of structural response signals corresponding to the current excitation source;

S3: The remaining piezoelectric sensors on the diagonal are used as excitation sources in turn, and step S2 is repeated respectively, and multiple pairs of structural response signals corresponding to different excitation sources are collected;

S4: Obtain the time difference of a pair of damage scattering signals corresponding to each pair of structural response signals; and determine the time difference by observing the respective damage scattering signals in each pair of structural response signals and qualitatively judging the order of succession positive and negative, and then determine the distance between the damage location and a pair of piezoelectric sensors as receivers.

S5: Use the hyperbolic positioning method to calculate the damage location and approximate range, including:

S51: Calculate the damage scattering distance difference according to the time difference obtained in step S4 and the group velocity c of the excitation signal propagating in the structure to be tested; obtain a hyperbolic trajectory according to the damage scattering distance difference, and combine all hyperbolic trajectories Calculate the damage location;

S52: Divide the monitoring area of the structure to be tested into micro-units, establish a coordinate matrix and a corresponding image matrix, and then perform pixel assignment on each element in the image matrix according to a probability density function to obtain a focused image of damage localization ; Then, according to the coordinates of each piezoelectric sensor of the excitation/sensing array and the damage position obtained in the step S51, the approximate range of the damage position is imaged to achieve accurate positioning and monitoring of the damage.

In the step S1, the diamond-shaped piezoelectric excitation/sensing array is composed of 4s piezoelectric sensors, and each side of the piezoelectric excitation/sensing array is respectively provided with s+1 equally spaced distributions The piezoelectric sensor, s is a positive integer.

In the step S1, the piezoelectric excitation/sensing array is composed of four piezoelectric sensors, and the piezoelectric excitation/sensing array is square.

In the step S2, the excitation signal of the single Lamb wave mode is loaded on the excitation source by an ultrasonic narrowband signal generated by a signal generator after passing through a power amplifier, and is excited by adjusting the frequency; the structure The response signal is acquired by collecting the signal received by the receiver and entering a computer through a digital oscilloscope.

In the step S2, the excitation signal is a sinusoidal signal or a sinusoidal signal modulated by a window function.

Preferably, the excitation signal is a sinusoidal signal modulated by a Hanning window.

The step S4 includes:

S41: Normalize all the structural response signals collected in steps S2 and S3 to eliminate errors caused by differences in the performance of each piezoelectric sensor;

S42: Subtract each pair of normalized structural response signals separately. According to the mid-perpendicular line theorem, after subtraction, the direct sensing signals in the structural response signals will be cancelled each other, thereby separating a pair of damage scattering signals. signal, and the time difference of the pair of damage scattering signals is determined according to the time points corresponding to the respective signal peaks of the pair of damage scattering signals.

Preferably, in the step S4, the time difference Δt _iAB of the pair of damage scattering signals is:

Δt _iAB =t _iA -t _iB ,

Among them, t _iA and t _iB represent the damage scattering time from the piezoelectric sensor P _i as the excitation source to the pair of piezoelectric sensors P _A and P _B as the receiver, respectively, i is the ordinal number of the piezoelectric sensor as the excitation source, A and B are respectively the ordinal numbers of the piezoelectric sensors as receivers; when i=1~4, if i=1 or 3, A=2, B=4, if i=2 or 4, A=1, B =3;

And in the step S4, when Δt _iAB is negative, the damage position is close to the A-th piezoelectric sensor P _A and away from the B-th piezoelectric sensor P _B ; when Δt _iAB is positive, the damage position is far away from the A-th piezoelectric sensor P B . The piezoelectric sensor P _A is close to the B-th piezoelectric sensor P _B . When Δt _iAB =0, the damage position is located at the mid-perpendicular line of the connection between the A-th piezoelectric sensor P _A and the B-th piezoelectric sensor P _B . , or there is no damage in the structure under test.

The non-reference Lamb wave damage monitoring method based on hyperbolic positioning of the present invention adopts the existing piezoelectric sensors to realize the rhombus array whose diagonal lines are perpendicular to each other, and selects the piezoelectric sensors located on a diagonal line as the excitation source to excite the For the excitation signal of a single Lamb wave mode, a pair of piezoelectric sensors on a straight line perpendicular to the diagonal are used as receivers to collect the structural response signal, which makes the array simple and the number is small, and the experimental measurement operation is simple. The principle of curve damage localization realizes the detection and monitoring of structural damage without the need for a reference signal. Therefore, it is not affected by changes in the structure and external conditions, which improves the accuracy of structural damage monitoring, and does not require a large number of The structure responds to signal data and complex signal processing and calculation steps, and has good operability for engineering applications. In addition, the Lamb wave damage monitoring method of the present invention realizes the separation of the damage scattering signal under active Lamb wave excitation and the direct sensing signal through signal normalization and signal difference, and highlights the damage scattering signal information.

Description of drawings

FIG. 1 is a schematic diagram of a piezoelectric excitation/sensing array used in a reference-free Lamb wave damage monitoring method based on hyperbolic positioning according to an embodiment of the present invention.

FIG. 2 is a schematic diagram of a piezoelectric excitation/sensing array used in a reference-free Lamb wave damage monitoring method based on hyperbolic positioning according to another embodiment of the present invention.

3A-3B are schematic diagrams of a selected excitation signal of a single Lamb wave mode of a reference-free Lamb wave damage monitoring method based on hyperbolic positioning according to an embodiment of the present invention, and FIG. 3A shows the excitation signal of a single Lamb wave mode Figure 3B shows the frequency domain waveform of the excitation signal of a single Lamb wave mode.

4A-4D are the structural response signals collected by the piezoelectric excitation/sensing array shown in The structural response signals collected when the four piezoelectric sensors are used as receivers. Figures 4C and 4D respectively show the structural response signals collected when the second piezoelectric sensor is used as the excitation source and the first and third piezoelectric sensors are used as receivers. .

Figures 5A-5B are respectively the damage scattering signals obtained by normalizing the structural response signals shown in Figures 4A-4D and then subtracting them in pairs, wherein Figure 5A shows the structures of the second and fourth piezoelectric sensors A pair of damage scattering signals obtained by first normalizing the response signals and then subtracting the difference. Figure 5B shows a pair of damage scattering signals obtained by first normalizing and then subtracting the structural response signals of the first and third piezoelectric sensors. .

FIG. 6 is a graph showing the results of imaging monitoring of damage localization in the structure to be tested.

Detailed ways

The present invention will be further described below with reference to specific embodiments. It should be understood that the following examples are only used to illustrate the present invention and not to limit the scope of the present invention.

The invention provides a reference-free Lamb wave damage monitoring method based on hyperbolic positioning, the method comprising the following steps:

Step S1: As shown in Figure 1, on a structure to be tested, according to the size of the required monitoring area, a rhombus (preferably a square, a square is a special rhombus) piezoelectric excitation/sensing array is arranged;

As shown in Figure 1, in this embodiment, the piezoelectric excitation/sensing array is composed of four piezoelectric sensors Pi ( _i =1, 2, 3, 4), and the structure to be tested is an aluminum plate structure with a size of 200 ×100×1mm ³ , taking the center of the plate structure as the coordinate origin, the coordinates of each piezoelectric sensor of the piezoelectric excitation/sensing array are respectively the first piezoelectric sensor P ₁ (100, 10), the second piezoelectric sensor P ₂ (60,50), third piezoelectric sensor P ₃ (100, 90), fourth piezoelectric sensor P ₄ (140, 50), in mm, first piezoelectric sensor P ₁ and third piezoelectric sensor P The diagonal line where ₃ is located is perpendicular to the diagonal line where the second piezoelectric sensor P ₂ and the fourth piezoelectric sensor P ₄ are located. The damage form in the experimental measurement is a typical through-hole damage, with a diameter of 5 mm and a center position of (80, 50) in mm.

In addition, in other embodiments, the piezoelectric excitation/sensing array may also be composed of 4s piezoelectric sensors, s is a positive integer, and each side (including the vertex) of the rhombus piezoelectric excitation/sensing array is on the There are respectively s+1 piezoelectric sensors distributed at equal intervals, for example, a piezoelectric excitation/sensing array composed of 12 piezoelectric sensors as shown in FIG. 2 . Therefore, the geometric feature of the piezoelectric excitation/sensing array is that the diagonal lines are perpendicular to each other, and the excitation and reception of ultrasonic Lamb waves are performed by using the forward and reverse piezoelectric effect of the piezoelectric sensor. In the experimental measurement, the piezoelectric sensor is directly pasted on the same side surface of the structure to be measured by the instant adhesive.

Step S2: Select a piezoelectric sensor Pi ( _i =1, 2, 3, 4) located on a diagonal line of the rhombus piezoelectric excitation/sensing array as an excitation source, and excite the structure to be tested The excitation signal of a single Lamb wave mode is obtained, and a pair of piezoelectric sensors P _A and P _B on a straight line perpendicular to the diagonal are selected as receivers, and a pair of structural response signals f corresponding to the current excitation source are collected respectively. _iA , f _iB , wherein the excitation signal of the single Lamb wave mode is preferably the excitation signal of a single S0 mode, i is the ordinal number of the piezoelectric sensor as the excitation source, A and B are the ordinal numbers of the piezoelectric sensor as the receiver, respectively . For example, as shown in Figure 1, the piezoelectric excitation/sensing array consists of 4 piezoelectric sensors, that is, s=1, i=1~4, if i=1 or 3, A=2, B=4 , if i=2 or 4, A=1, B=3; in addition, as shown in Figure 2, the piezoelectric excitation/sensing array consists of 12 piezoelectric sensors, that is, s=3, i=1, 4 ,7,10, if i=1 or 7, A=2, B=12; A=3, B=11; A=4, B=10; A=5, B=9; and A=6, B=8; if i=4 or 10, A=3, B=5; A=2, B=6; A=1, B=7; A=12, B=8; and A=11, B= 9.

In this embodiment, the excitation signal is a 5-period sinusoidal signal modulated by a Hanning window, the center frequency is 250 kHz, and the time-domain waveform and frequency-domain waveform are shown in FIG. 3A and FIG. 3B , respectively. In addition, the excitation signal may also be a sinusoidal signal modulated by other window functions, such as a Heaviside step function, or a sinusoidal signal may be directly used. However, in general, the excitation signal is a sinusoidal signal modulated by a Hanning window, which makes the signal energy more concentrated and the waveform change more stable.

For example, as shown in FIG. 4A and FIG. 4B , when i=1, that is, when the first piezoelectric sensor P ₁ is used as the excitation source, and the second piezoelectric sensor P ₂ and the fourth piezoelectric sensor P ₄ are respectively used as receivers, A time domain diagram of a pair of acquired structural response signals f ₁₂ and f ₁₄ . The structural response signals f ₁₂ and f ₁₄ can be divided into direct sensing signals and damage scattering signals, wherein the direct sensing signal is the first signal with a relatively large wave packet peak and a relatively clear waveform, and the damage scattering signal is In the latter wave packet peak value is relatively small, as shown in Figure 4B, the direct sensing signal and the damage scattering signal are clearly separated, but sometimes the two overlap each other. The direct sensing signals in the pair of structural response signals f ₁₂ and f ₁₄ shown in Fig. 4A and Fig. 4B are basically the same, while the damage scattering signals are significantly different, which indicates that there is damage in the structure to be tested, and the damage location is not It is located on the vertical line of the diagonal where the second piezoelectric sensor P ₂ and the fourth piezoelectric sensor P ₄ are located; as shown in FIG. 4C and FIG. 4D , when i=2, the second piezoelectric sensor P ₂ is used as the When the excitation source, the first piezoelectric sensor P ₁ and the third piezoelectric sensor P ₃ are used as receivers, respectively, the time domain diagram of a pair of structural response signals f ₂₁ and f ₂₃ is collected, the pair of structural response signals f ₂₁ and _f23 , there is no obvious difference, indicating that there is no damage in the structure to be tested or the damage position is located on the vertical line of the diagonal line where the _first piezoelectric sensor P1 and the _third piezoelectric sensor P3 are located. At the same time, according to the previous pair of structures In response to the determination of the signals f ₁₂ and f ₁₄ , it can be preliminarily determined that the damaged position is located on the mid-perpendicular line of the diagonal where the first piezoelectric sensor P ₁ and the third piezoelectric sensor P ₃ are located.

Thus, the piezoelectric sensor P _i as an excitation source and a pair of piezoelectric sensors P _A and P _B as receivers are constructed as a set of monitoring channels. The hardware part used in the detection method of the present invention is the same as the hardware part of the monitoring system of the traditional method, and generally consists of the following parts: a computer, a signal generator, a power amplifier, a digital oscilloscope and various connecting lines. Wherein, the excitation signal of the single Lamb wave mode is loaded on the excitation source by an ultrasonic narrowband signal generated by a signal generator after passing through a power amplifier, and is excited by adjusting the frequency; the structural response signal is received by receiving The signal received by the receiver is collected by a digital oscilloscope and entered into a computer to be collected separately. The sampling frequency is set to 25MHz.

In addition, the step S2 also includes: acquiring the time difference between the direct sensing signal and the excitation signal in the structural response signals f _iA , f _iB , and then according to the linear distance between the excitation source and the receiver (known) , calculate the group velocity c of the excitation signal propagating in the structure to be tested, and assume that the group velocity does not change before and after the damage.

Step S3: The remaining piezoelectric sensors on the diagonal are used as excitation sources in turn, and step S2 is repeated respectively to collect multiple pairs of structural response signals corresponding to different excitation sources, thereby obtaining the structural response signals of the entire monitoring area.

Step S4: obtaining the time difference of a pair of damage scattering signals corresponding to each pair of structural response signals, specifically including:

Step S41: normalize all the structural response signals collected in steps S2 and S3 to eliminate errors caused by differences in the performance of each piezoelectric sensor;

The normalization process is carried out using the following formula:

Among them, x(t) is the structural response signal, Max is the maximum value of the structural response signal x(t), and y(t) is the normalized structural response signal.

Step S42 : subtracting each pair of normalized structural response signals respectively, according to the mid-perpendicular line theorem, the direct sensing signals in the structural response signals will cancel each other after the subtraction, so as to separate a pair of damages. and determine the time difference Δt _iAB of the pair of damage scattering signals according to the time points corresponding to the respective signal peaks of the pair of damage scattering signals; that is, Δt _iAB =t _iA -t _iB , where t _iA and t _iB are respectively represents the damage scattering time from the piezoelectric sensor P ₁ as the excitation source to a pair of piezoelectric sensors P _A and P _B as the receiver (including the A-th piezoelectric sensor P _A and the B-th piezoelectric sensor P _B ), i is the ordinal number of the piezoelectric sensor as the excitation source, A and B are the ordinal numbers of the piezoelectric sensor as the receiver, respectively, when i=1~4, if i=1 or 3, A=2, B=4, if i =

2 or 4, A=1, B=3.

For example, as shown in FIG _{. 5A} , the structure response signal ( _ie i =1, A=2, B=4), after normalization, the difference and subtraction can eliminate the direct sensing signal to obtain a pair of damage scattering signals, which can be determined according to the time points corresponding to the respective signal peaks of the pair of damage scattering signals The time difference Δt _iAB of a pair of damage scattering signals; as shown in FIG. 5B , collected when the second piezoelectric sensor P ₂ is used as the excitation source, and the first piezoelectric sensor P ₁ and the third piezoelectric sensor P ₃ are respectively used as receivers To the structural response signal (i.e. i=2, A=1, B=3), the difference and subtraction after normalization can eliminate the direct sensing signal, and it is found that the signal tends to be stable and fluctuates at the 0 position, which can be preliminarily determined. The time difference Δt _iAB = 0 for the damage scattering signal, and the damage position is located on the vertical line of the diagonal where the first piezoelectric sensor P ₁ and the third piezoelectric sensor P ₃ are located.

In addition, the step S4 further includes: by respectively observing the respective damage scattering signals in each pair of structural response signals, and qualitatively judging the order of succession, thereby determining the time difference Δt _iAB of the pair of damage scattering signals in the step S4 positive and negative, and then determine the distance of the damage location from the piezoelectric sensors P _A and P _B as receivers. In this way, it can be determined which of the two symmetrical hyperbolas the damage position is specifically on, and it can be determined that the damage position is specifically on a hyperbolic trajectory, and the accurate damage position can be obtained by eliminating the artifacts.

Wherein, when the time difference Δt _iAB of a pair of damage scattering signals in the step S4 is negative, it means that the damage position is close to the A-th piezoelectric sensor P _A and far away from the B-th piezoelectric sensor P _B ; When the time difference Δt _iAB is positive, it means that the damage position is far away from the A-th piezoelectric sensor P _A and close to the B-th piezoelectric sensor P _B . When the time difference Δt _iAB of the damage scattering signal is 0, it means that the damage position is located in the A-th piezoelectric sensor. On the vertical line of the connection between the electrical sensor P _A and the B-th piezoelectric sensor P _B , or there is no damage in the structure to be measured, where i is the serial number of the piezoelectric sensor as the excitation source, A and B are the The ordinal number of the piezoelectric sensor of the receiver, when i=1 to 4, if i=1 or 3, A=2, B=4, if i=2 or 3, A=1, B=3.

For example, as shown in Figure 4A, the damage scattering signal is not obvious, indicating that the damage scattering signal overlaps with the direct sensing signal; as shown in Figure 4B, the damage scattering signal is obvious, indicating that the damage scattering signal is behind the direct sensing signal; Reaching the second piezoelectric sensor P ₂ , and then reaching the fourth piezoelectric sensor P ₄ , it is determined that the time difference Δt _iAB is less than 0 (i=1, A=2, B=4); and it can be known that the time difference Δt ₁₂₄ is less than 0, The damaged position is close to the second piezoelectric sensor P ₂ and far away from the fourth piezoelectric sensor P ₄ . In addition, it can be seen from the above description about FIG. 5B that the time difference between a pair of damage scattering signals Δt _iAB = 0, and it shows that the damage position is located in the diagonal line where the first piezoelectric sensor P ₁ and the third piezoelectric sensor P ₃ are located. on the vertical line.

Step S5: Using the hyperbolic positioning method, the position and approximate range of the damage are calculated, and the health of the structure to be tested is analyzed and judged.

The step S5 specifically includes:

Step S51: Calculate the damage scattering distance difference Δd _iAB according to the time difference Δt _iAB of the pair of damage scattering signals in the step S4 and the group velocity c of the excitation signal in the step S2 propagating in the structure to be tested , that is, Δd _iAB =c×Δt _iAB ; multiple hyperbolic trajectories are obtained according to multiple damage scattering distance differences Δd _iAB , and the damage positions are calculated by combining all hyperbolic trajectories.

Specifically, according to the knowledge of geometry, the set of points whose distances from a pair of fixed points are a constant value is a hyperbola, so that each damage scattering distance difference Δd _iAB and its corresponding pair of piezoelectric sensors P can be determined as a hyperbola. _A pair of hyperbolic trajectories are obtained from A and P _B , and multiple pairs of hyperbolic trajectories are obtained for multiple damage scattering distance differences Δd _iAB . The position with the most intersections of all hyperbolic trajectories can be determined as the damage position.

Thus, in step S51, the damage position of the structure to be tested has been determined from the mathematical geometric formula.

Step S52: Divide the monitoring area of the structure under test into tiny cells, and establish a coordinate matrix and a corresponding image matrix, so that each element in the image matrix represents a tiny cell in the structure under test, and then according to a probability The density function assigns a pixel value to each element in the image matrix to obtain a focused image of damage location; then, according to the coordinates of each piezoelectric sensor in the excitation/sensing array and the damage position obtained in the step S51, the damage position is determined. Perform localization imaging in a rough range to achieve accurate localization and monitoring of damage.

Here, the damage scattering distance difference Δd _iAB is to obtain a hyperbolic trajectory in the sense of mathematical geometry, and the probability density function represents the probability density that the coordinate point is the damage location. The closer it is to the hyperbolic trajectory, the greater the probability value indicating that the coordinate point is the damage location, and the probability value is used as the pixel value of positioning imaging. In this way, pixel assignment to the image (represented by a two-dimensional matrix) is realized. Different pixel values are represented by different colors. The darker the color, the greater the probability of damage.

Wherein, the probability density function and the cumulative distribution function adopt the following expressions:

Correspondingly, the lesion localization focused image is:

D _IJ (x _m , y _n )=1-[F(z _IJ )-F(-z _IJ )]

Among them, z _IJ is the shortest distance from the tiny unit to the hyperbolic trajectory determined by the pair of piezoelectric sensors I, J as the receiver, the unit is mm, where I, J represent the ordinal number of the pair of piezoelectric sensors as the receiver , for example, the first piezoelectric sensor P ₁ is used as the excitation source, the second piezoelectric sensor P ₂ and the fourth piezoelectric sensor P ₄ are used as the receiver, then I, J are expressed as 2, 4 respectively; F(z _IJ ) is Cumulative distribution function, f(z) is the probability density function, D _IJ (x _m , y _n ) is the pixel value of any element in the image matrix, x _m , y _n are the position of the point in the image matrix, the unit is mm, which can reflect the coordinates of the corresponding micro-unit, DI _total represents the focused image of damage location obtained by all piezoelectric sensors, and N is the number of piezoelectric sensors.

In addition, as shown in FIG. 6 , the step S52 further includes: after obtaining the focused image of damage location, setting a threshold for the pixel values of all elements in the image matrix to display the damage location more clearly and accurately. The threshold set in the legend is 90%, which means that more than 90% of the maximum value is displayed, and 90% less than the maximum value does not need to be displayed. The positioning results are displayed near the actual damage location, and the approximate range of the damage location can also be described. The monitoring results show that the method of the present invention basically detects the damage of the structure to be measured without the need for a reference signal, and realizes the location monitoring without reference damage.

The above descriptions are only preferred embodiments of the present invention, and are not intended to limit the scope of the present invention. Various changes can be made to the above-mentioned embodiments of the present invention. All simple and equivalent changes and modifications made according to the claims and descriptions of the present application shall fall within the protection scope of the claims of the present invention. What is not described in detail in the present invention is conventional technical content.

Claims (6)

1. a non-reference Lamb wave damage monitoring method based on hyperbolic positioning, is characterized in that, comprises: Step S1: On a structure to be tested, according to the size of the required monitoring area, a diamond-shaped piezoelectric excitation/sensing array is arranged; Step S2: Select a piezoelectric sensor located on a diagonal line of the rhombus piezoelectric excitation/sensing array as the excitation source, excite the excitation signal of a single Lamb wave mode in the structure to be measured, and select the excitation signal corresponding to the A pair of piezoelectric sensors on a straight line perpendicular to the diagonal are used as receivers, respectively collecting a pair of structural response signals corresponding to the current excitation source; Step S3: successively use the remaining piezoelectric sensors on the diagonal as excitation sources, repeat step S2 respectively, and collect multiple pairs of structural response signals corresponding to different excitation sources; Step S4 : obtaining the time difference of a pair of damage scattering signals corresponding to each pair of structural response signals; by separately observing the respective damage scattering signals in each pair of structural response signals, and qualitatively judging the sequence in order, thereby determining the time difference positive and negative, and then determine the distance of the damage location from a pair of piezoelectric sensors as receivers; Step S5: Use the hyperbolic positioning method to calculate the damage location and approximate range, including: Step S51: Calculate the damage scattering distance difference according to the time difference obtained in step S4 and through the group velocity c of the excitation signal propagating in the structure to be tested; obtain a hyperbolic trajectory according to the damage scattering distance difference, and combine all hyperbolas Trajectory calculation to get the damage location; Step S52: Divide the monitoring area of the structure to be tested into micro-units, establish a coordinate matrix and a corresponding image matrix, and then perform pixel assignment on each element in the image matrix according to a probability density function to obtain a damage localization focus. image; then, according to the coordinates of each piezoelectric sensor of the excitation/sensing array, and the damage position obtained in the step S51, the approximate range of the damage position is imaged. 2. The non-reference Lamb wave damage monitoring method based on hyperbolic positioning according to claim 1, wherein in the step S1, the rhombus piezoelectric excitation/sensing array is composed of 4s piezoelectric sensors Each side of the piezoelectric excitation/sensing array is respectively provided with s+1 piezoelectric sensors distributed at equal intervals, and s is a positive integer. 3. The non-reference Lamb wave damage monitoring method based on hyperbolic positioning according to claim 2, wherein in the step S1, the piezoelectric excitation/sensing array is composed of 4 piezoelectric sensors, And the piezoelectric excitation/sensing array is square. 4. the non-reference Lamb wave damage monitoring method based on hyperbolic positioning according to claim 1, is characterized in that, in described step S2, the excitation signal of described single Lamb wave mode is generated by a signal generator. The ultrasonic narrowband signal is loaded onto the excitation source after passing through a power amplifier, and is excited by adjusting the frequency; the structural response signal is acquired by collecting the signal received by the receiver and entering a computer through a digital oscilloscope. 5. The non-reference Lamb wave damage monitoring method based on hyperbolic positioning according to claim 1, wherein the step S4 comprises: Step S41: normalize all the structural response signals collected in steps S2 and S3 to eliminate errors caused by differences in the performance of each piezoelectric sensor; Step S42 : subtracting each pair of normalized structural response signals respectively, according to the mid-perpendicular line theorem, the direct sensing signals in the structural response signals will cancel each other after the subtraction, so as to separate a pair of damages. The scatter signal is obtained, and the time difference of the pair of damage scatter signals is determined according to the time points corresponding to the respective signal peaks of the pair of damage scatter signals. 6 . The reference-free Lamb wave damage monitoring method based on hyperbolic positioning according to claim 1 , wherein, in the step S4 , the time difference Δt _iAB of the pair of damage scattering signals is: 6 . Δt _iAB =t _iA -t _iB , Among them, t _iA and t _iB represent the damage scattering time from the piezoelectric sensor P _i as the excitation source to the pair of piezoelectric sensors P _A and P _B as the receiver, respectively, i is the ordinal number of the piezoelectric sensor as the excitation source, A and B are respectively the ordinal numbers of the piezoelectric sensors as receivers; when i=1~4, if i=1 or 3, A=2, B=4, if i=2 or 4, A=1, B =3; And in the step S4, when Δt _iAB is negative, the damage position is close to the A-th piezoelectric sensor P _A and away from the B-th piezoelectric sensor P _B ; when Δt _iAB is positive, the damage position is far away from the A-th piezoelectric sensor P B . The piezoelectric sensor P _A is close to the B-th piezoelectric sensor P _B . When Δt _iAB =0, the damage position is located at the mid-perpendicular line of the connection between the A-th piezoelectric sensor P _A and the B-th piezoelectric sensor P _B . , or there is no damage in the structure under test.

Images