CN104792865A - Recognizing and positioning method of small defects of pipelines through ultrasonic guided waves based on fractal dimensions - Google Patents

Abstract

The invention discloses a recognizing and positioning method of small defects of pipelines through ultrasonic guided waves based on fractal dimensions. The recognizing and positioning method comprises the following steps: stimulating signals of ultrasonic guide waves through a waveform generator, a power amplifier and a piezoelectric ring, receiving actual measurement signals through a piezoelectric plate, and recording a time travel curve of the ultrasonic guide waves spreading in a pipeline through a digital oscilloscope; establishing a signal detecting system of a Duffing oscillator; confirming the F value which can be used for recognizing the signals of ultrasonic guide waves according to the change of fractal dimensions with the driving amplitude F; defining a moving window function, scanning the actual measurement signals through the moving window function, calculating the fractal dimension of each section of signals, if the fractal dimension is equal to 2, showing that the pipeline is complete, if the pipeline is larger than 2, showing that the pipeline has defects, and positioning the defects by utilizing the peak value of an envelope curve of the fractal dimension and the window moving speed (tao is equal to 1 microseconds). Through the adoption of the recognizing and positioning method disclosed by the invention, small defects of different damage degrees in the pipelines can be effectively positioned, so that the sensitivity of ultrasonic guide waves recognizing the small defects is improved, and the detection range is effectively enlarged.

Description

Based on the little defect recognition of pipe ultrasonic guided wave and the localization method of dimension

Technical field

The present invention relates to a kind of ultrasonic guided wave detection technology, especially a kind of little defect recognition of pipe ultrasonic guided wave based on dimension and localization method, belong to technical field of nondestructive testing.

Background technology

In recent years, pipeline is widely used in industry-by-industry, has become the fifth-largest means of transport after railway, highway, air transportation, water transport.But, because pipeline accident takes place frequently, health monitoring is carried out to pipeline and becomes very necessary.In recent years, ultrasonic guided wave detection technology becomes long distance line detection innovative techniques.Compared with conventional ultrasonic wave detection technique, supersonic guide-wave excites along waveguiding structure length direction, and its sensing range is " line " but not " point ", and sensing range can reach 50 ~ 100 meters, has become the important method that long distance line detects.

At present, detection technique for supersonic guide-wave has a lot, mainly three parts can be divided into: one, the Propagation Characteristics of supersonic guide-wave in pipeline configuration, as the Dispersion research in different pipeline feature and under different service condition, attenuation Characteristics, propagation mode and MODAL TRANSFORMATION OF A research etc.; Two, the exciting and receiving trap research of supersonic guide-wave, as piezoelectric effect, magnetostrictive effect and based on pulse laser formula guided wave sensor design and applied research etc.; Three, signal analysis and defect characteristic extract, as wavelet transformation, and quick FFT, defect parameters identification etc.Pertinent literature shows to have abundant achievement in research in above-mentioned three.But in the commercial Application of ultrasonic guided wave detection technology, still there are some problems.Especially for actual pipeline such as some liquid-filling pipe, buried pipelines etc., excite, due to these pipeline features, even make the signal received also mostly be difficult to see obvious flaw echoes under larger defect, also have some long distances, the echoed signal under little defect is also difficult to be observed, therefore, domestic and international researcher more and more payes attention to analysis to ultrasonic guided wave signals and research, and has developed many methods effectively.Chaos system, owing to having initial value sensitivity, if be input in system using weak signal as the initial value of chaos system, effectively can identify weak signal, and effectively reduce signal-noise ratio threshold, improve detection sensitivity from the response of chaos system.Therefore, the attention of researchist is caused based on the Technique of Weak Signal Detection of chaos system.

Conventional chaos detection system mainly contains Lorenz system and Duffing system.There are many researchers based on off-resonance parametric excitation chaos suppression principle, utilize controlled Lorenz system to realize the detection of weak harmonic signal under strong noise background.But in recent years, the weak periodic signal detection method based on Duffing oscillator obtains further investigation.Be carried out by doctor Donald.L.Birx of Dayton university of the U.S. in 1992 at first, goed deep into theoretical research owing to lacking about basic theory is not perfect at that time.Subsequently, researcher utilizes and realizes detecting to the parameter resonance perturbation of Non-Self-Governing Du Fen chaos system, and the correlative study such as system statistics characteristic, weak signal cycle detection, unknown frequency detection, signal amplitude estimation carried out based on chaos detection technology, but major part is round cycle and harmonic signal.Now increasing scholar is also had to utilize Du Fen chaos system to carry out the identification of weak ultrasonic guided wave signals.But mostly concentrate on the identification utilizing qualitative index phase path figure to guided wave signals, and further little defect tracking location is seemed more urgent.

Summary of the invention

The object of the invention is the defect in order to solve conventional ultrasonic wave detection technique, a kind of little defect recognition of pipe ultrasonic guided wave based on dimension and localization method are provided, the method can carry out effective location to the little defect of ducted Injured level, thus improve the little defect of supersonic guide-wave identification sensitivity, effectively extend sensing range.

Object of the present invention can reach by taking following technical scheme:

Based on the little defect recognition of pipe ultrasonic guided wave and the localization method of dimension, comprise the following steps:

1) waveform generator generates the signal through the modulation of Hanning window, is amplified, remake the piezoelectric ring for pipeline one side end face by power amplifier, excites the ultrasonic guided wave signals through the modulation of Hanning window, makes supersonic guide-wave travel through all positions of pipeline;

2) receive measured signal by being arranged on piezoelectric patches on pipeline, and by the time-history curves that digital oscilloscope record supersonic guide-wave is propagated in the duct, by digital oscilloscope by the information transmission of record to computing machine;

3) set the incident wave signal of the intact pipeline collected in advance and the pure noise signal of numerical simulation as detection signal, according to centre frequency, the sample frequency of this detection signal, and duffing equation characteristic, structure Duffing vibrator signal detection system;

4) the incident wave signal of the intact pipeline collected in advance and the pure noise signal of numerical simulation are input to Duffing vibrator signal detection system respectively, by compare no signal input, incident wave signal and pure noise signal time, dimension, with the change of driving force amplitude F, determines the F value that can be used for identifying ultrasonic guided wave signals;

5) Moving Window function is defined, select window length 2 δ=50 μ s, window translational speed τ=5 μ s, by Moving Window function scanning measured signal, calculate the dimension of each segment signal, if the dimension between incident wave and edge echo equals 2, then pipeline is intact; If the dimension between incident wave and edge echo is greater than 2, then pipeline defectiveness, enters step 6);

6) in the moment utilizing the envelope peak value determination incident wave of dimension, edge echo and flaw echo to receive, according to the time scale relation between three, defect of pipeline location is carried out;

7) window translational speed τ=1 μ s is utilized accurately to locate ducted little defect.

As a kind of embodiment, step 1) the described ultrasonic guided wave signals expression formula through the modulation of Hanning window is as follows:

$s\left(t\right)=\left[\frac{1}{2}(1-\cos \frac{w_{c}t}{n})\right].\sin \left(w_{c}t\right)---\left(1\right)$

Wherein, n is the single audio frequency number selected, w _c=2 π f _c, f _cfor the centre frequency of measured signal, be 70KHz.

As a kind of embodiment, step 3) described structure Duffing vibrator signal detection system, specific as follows:

A) choose duffing equation, it comprises the complex state of vibration, fork, chaos, and its expression formula is:

$\stackrel{\·\·}{x}+k\stackrel{\·}{x}-x^3+x^5=F\cos \mathrm{\ωt}---\left(2\right)$

Wherein, k is damping ratio, (-x ³+ x ⁵) be nonlinear resilience item; Fcos ω t is driving force item, and F is driving force amplitude, and ω is driving force angular frequency;

B) set the incident wave signal of the intact pipeline collected in advance and the pure noise signal of numerical simulation as detection signal and its centre frequency is 70KHz, sample frequency is 50M time/second, then sets the ω ≈ 0.439823rad/ μ s of duffing equation, integration step h=0.02 μ s, damping ratio k=0.5, improves as follows to formula (2):

$\stackrel{\·\·}{x}+k\stackrel{\·}{x}-x^3+x^5=F\cos \mathrm{\ωt}+\stackrel{\‾}{s}\left(t\right)---\left(3\right)$

As a kind of embodiment, step 5) definition of described dimension is specific as follows:

A) continuous dynamical system x=F (x) is tieed up, in the t=0 moment, with x for n ₀centered by, || δ ₀(x ₀, 0) || for radius does the sphere of a n dimension; Along with the evolution of time, be namely deformed into the ellipsoid of n dimension at this sphere of t, if half axial length of i-th of this ellipsoid change in coordinate axis direction is, then this system i-th Lyapunov index is:

${\mathrm{\λ}}_i=\underset{t\→\∞}{\lim }\frac{1}{t}\ln \frac{\left|\right|\mathrm{\δ}{x}_i(x_0,t)\left|\right|}{\left|\right|\mathrm{\δ}{x}_i(x_0,0)\left|\right|}---\left(4\right)$

N maintains system, corresponding n Lyapunov exponential quantity, as long as there is a Lyapunov index to be greater than 0, is in chaos state with regard to illustrative system, two-dimentional Non-Self-Governing Duffing vibrator signal detection system formula (3) represented is rewritten as three-dimensional autonomous system, as shown in the formula:

$\left.\begin{array}{c}\stackrel{\·}{x}=v\\ \stackrel{\·}{v}=-\mathrm{kv}+x^3-x^5+F\cos \left(\mathrm{\ωz}\right)+\stackrel{\‾}{s}\left(z\right)\\ \stackrel{\·}{z}=1\end{array}\right\}---\left(5\right)$

Formula (4) is utilized to calculate three Lyapunov index λ corresponding to Duffing vibrator signal detection system respectively ₁, λ ₂and λ ₃;

B) according to the above-mentioned Lyapunov index that will calculate, a kind of definition of dimension is obtained, as follows:

$D=j+\frac{\underset{i=1}{\overset{j}{\mathrm{\Σ}}}{\mathrm{\λ}}_i}{\left|{\mathrm{\λ}}_{j+1}\right|}---\left(6\right)$

Wherein, meet the dimension that delimiting period moves and quasi-periodic motion is corresponding is again 2.

As a kind of embodiment, step 5) described Moving Window function is defined as follows:

S ^*＝g(t-nτ)S (7)

$g(t-\mathrm{n\τ})=1,t\∈(\mathrm{n\τ}-\mathrm{\δ},\mathrm{n\τ}+\mathrm{\δ}),n=\frac{\mathrm{\δ}}{\mathrm{\τ}},\frac{\mathrm{\δ}+\mathrm{\τ}}{\mathrm{\τ}},\frac{\mathrm{\δ}+2\mathrm{\τ}}{\mathrm{\τ}}...,\frac{N-\mathrm{\δ}}{\mathrm{\τ}}---\left(8\right)$

Wherein, S represents the full time-domain signal recorded, S ^*represent intercept signal, N represents signal length, and 2 δ represent window length, and τ represents window translational speed.

As a kind of embodiment, step 6) described in carry out defect of pipeline location and adopt following formula to calculate:

$d_x=\frac{t_2-t_1}{t_3-t_1}d---\left(9\right)$

Wherein, d _xrepresent that in pipe, defect is from the distance exciting end, d represents duct length, t ₁, t ₂and t ₃represent the moment that incident wave, flaw echo and edge echo receive respectively.

As a kind of embodiment, step 7) the described window translational speed τ=1 μ s that utilizes accurately locates ducted little defect, specific as follows:

When first utilizing dimension peak value to provide window translational speed τ=1 μ s, the echo moment in centre position, then according to the velocity of propagation that dispersion curve calculates, in conjunction with following formula, ducted little defect is accurately located:

2d＝ct (10)

Wherein, c is the velocity of propagation calculated according to dispersion curve, and t represents the moment of the ripple received, and d represents that t pipeline location is from the distance exciting end.

The present invention has following beneficial effect relative to prior art:

1, the present invention is the little defect recognition of pipe ultrasonic guided wave based on dimension and localization method, can improve little flaw detection sensitivity, and the little defective locations of effective location, there is innovative significance, have a wide range of applications.

2, when the present invention detects the Weak Ultrasonic guided wave signals that ducted little defect reflection obtains, force value is instigated by what select Duffing vibrator signal detection system, then according to the size of the dimension calculated, just can find out whether system enters chaos state easily, thus whether defectiveness echoed signal is described, and then complete defect recognition.

3, the present invention gives the defect moment according to the envelope peak value of dimension, and then can be regarded as out defective locations, complete defect location, can also, when selecting the Moving Window function scanning overall signal of window length 2 δ=50 μ s, window translational speed τ=5 μ s, window translational speed τ=1 μ s be utilized accurately to locate ducted little defect.

Accompanying drawing explanation

The principle of instrument block diagram that Fig. 1 adopts for the embodiment of the present invention 1.

Fig. 2 is the little defect recognition of pipe ultrasonic guided wave and the localization method process flow diagram of the embodiment of the present invention 1.

Fig. 3 is that the pure noise signal that the intact pipeline incidence of the embodiment of the present invention 2 involves numerical simulation inputs the effect diagram that before and after Duffing vibrator signal detection system, dimension changes with driving force F.

Fig. 4 is the Moving Window function schematic diagram of the embodiment of the present invention 2.

Fig. 5 a is dimension and the time-history curves schematic diagram of the operating mode one of the embodiment of the present invention 2; The dimension of the operating mode two of Fig. 5 b embodiment of the present invention 2 and time-history curves schematic diagram; Fig. 5 c is dimension and the time-history curves schematic diagram of the operating mode three of the embodiment of the present invention 2; Fig. 5 d is dimension and the time-history curves schematic diagram of the operating mode four of the embodiment of the present invention 2.

Fig. 6 is the curve map that error r changes with 2 δ.

Fig. 7 is the change curve of dimension with signal length of the embodiment of the present invention 2.

Fig. 8 is the dispersion curve figure of the embodiment of the present invention 2.

Fig. 9 a is the positioning result figure of the operating mode two different windows translational speed of the embodiment of the present invention 2; Fig. 9 b is the positioning result figure of the operating mode three different windows translational speed of the embodiment of the present invention 2.

Embodiment

Embodiment 1:

As shown in Figure 1, the instrument that the present embodiment adopts comprises waveform generator, power amplifier, digital oscilloscope, piezoelectric ring and piezoelectric patches, described piezoelectric ring is arranged on pipeline one side end face, described piezoelectric patches is arranged on pipeline, described waveform generator, power amplifier are connected successively with piezoelectric ring, described waveform generator is connected with digital oscilloscope respectively with piezoelectric patches, and described digital oscilloscope is connected with computing machine.

As described in Figure 2, the little defect recognition of pipe ultrasonic guided wave of the present embodiment and localization method comprise the following steps:

1) waveform generator generates the signal through the modulation of Hanning window, is amplified, remake the piezoelectric ring for pipeline one side end face by power amplifier, excites the ultrasonic guided wave signals through the modulation of Hanning window, makes supersonic guide-wave travel through all positions of pipeline;

2) receive measured signal by piezoelectric patches on pipeline, and by the time-history curves that digital oscilloscope record supersonic guide-wave is propagated in the duct, by digital oscilloscope by the information transmission of record to computing machine;

3) set the incident wave signal of the intact pipeline collected in advance and the pure noise signal of numerical simulation as detection signal, according to centre frequency, the sample frequency of this detection signal, and duffing equation characteristic, structure Duffing vibrator signal detection system;

4) the incident wave signal of the intact pipeline collected in advance and the pure noise signal of numerical simulation are input to Duffing vibrator signal detection system respectively, by compare no signal input, incident wave signal and pure noise signal time, dimension, with the change of driving force amplitude F, determines the F value that can be used for identifying ultrasonic guided wave signals;

5) Moving Window function is defined, select window length 2 δ=50 μ s, window translational speed τ=5 μ s, by Moving Window function scanning measured signal, calculate the dimension of each segment signal, if the dimension between incident wave and edge echo equals 2, then pipeline is intact; If the dimension between incident wave and edge echo is greater than 2, then pipeline defectiveness, enters step 6);

6) in the moment utilizing the envelope peak value determination incident wave of dimension, edge echo and flaw echo to receive, according to the time scale relation between three, defect of pipeline location is carried out;

7) utilize window to move little speed (τ=1 μ s) accurately to locate ducted little defect.

Embodiment 2:

The present embodiment is for specific experiment, and carry out the little defect recognition of supersonic guide-wave and location in laboratory to pipeline, specific implementation process is as follows:

1) steel pipe of selection for pipeline 3mm × 50.75mm × 2.32mm, described piezoelectric ring and piezoelectric patches all adopt PZT5 material, wherein the size of piezoelectric ring is according to pipeline section manufacture, piezoelectric patches is of a size of 15.4mm × 3.2.mm × 0.9mm, use four install pipelines, four kinds of operating modes respectively, wherein operating mode one is flawless intact pipeline, and operating mode two, three and four is all utilize saw bow to excite the 1.5m place of end to arrange artificial defect at distance piezoelectric patches place, and design parameter is as shown in table 1 below.

Table 1 four kinds of operating mode defects arrange table

2) waveform generator generates the signal through the modulation of Hanning window, amplified by power amplifier, remake the piezoelectric ring of pipeline one side end face for four kinds of operating modes, excite the ultrasonic guided wave signals through the modulation of Hanning window, make supersonic guide-wave travel through all positions of pipeline;

Expression formula through the ultrasonic guided wave signals of Hanning window modulation is as follows:

$s\left(t\right)=\left[\frac{1}{2}(1-\cos \frac{w_{c}t}{n})\right].\sin \left(w_{c}t\right)---\left(1\right)$

Wherein, n is the single audio frequency number selected, w _c=2 π f _c, f _cfor the centre frequency of measured signal, be 70KHz.

3) received the measured signal of four kinds of operating modes by the piezoelectric patches on pipeline, and by the time-history curves that digital oscilloscope record supersonic guide-wave is propagated in the duct, by digital oscilloscope by the information transmission of record to computing machine;

4) Duffing vibrator signal detection system is constructed

A) choose duffing equation, duffing equation is the common model in Detection of Weak Signals, and the nonlinear system described by it shows multiple nonlinear characteristic, comprises the complex state of vibration, fork, chaos, and its expression formula is:

$\stackrel{\·\·}{x}+k\stackrel{\·}{x}-x^3+x^5=F\cos \mathrm{\ωt}---\left(2\right)$

Wherein, k is damping ratio, (-x ³+ x ⁵) be nonlinear resilience item; Fcos ω t is driving force item, and F is driving force amplitude, and ω is driving force angular frequency;

The incident wave signal of the intact pipeline b) setting laboratory to collect and the pure noise signal of numerical simulation are as detection signal and its centre frequency is 70KHz, sample frequency is 50M time/second, then sets the ω ≈ 0.439823rad/ μ s of duffing equation, integration step h=0.02 μ s, damping ratio k=0.5, improves as follows to formula (2):

$\stackrel{\·\·}{x}+k\stackrel{\·}{x}-x^3+x^5=F\cos \mathrm{\ωt}+\stackrel{\‾}{s}\left(t\right)---\left(3\right)$

So far, Duffing vibrator signal detection system construction complete.Utilize simple triangular transformation, formula (3) still can be classified as the form of formula (2) by abbreviation.Therefore, can think and be equivalent to input synperiodic sine (or cosine) signal amplitude and the phase place of the formula that changes (2) driving force item, cause the change of system output characteristic, thus realize input signal analysis.

5) dimension is defined

A) calculating of Lyapunov index

The essential characteristic of chaos system is that motion is very responsive to initial condition, the track exponentially mode of passing in time that two close initial values produce is separated, Lyapunov index is the quantitative target describing this phenomenon, the average index rate that As time goes on system that characterizes restrains between adjacent orbit or disperse in phase space.

Continuous dynamical system x=F (x) is tieed up, in the t=0 moment, with x for n ₀centered by, || δ x (x ₀, 0) || for radius does the sphere of a n dimension; Along with the evolution of time, be namely deformed into the ellipsoid of n dimension at this sphere of t, if half axial length of i-th of this ellipsoid change in coordinate axis direction is, then this system i-th Lyapunov index is:

${\mathrm{\λ}}_i=\underset{t\→\∞}{\lim }\frac{1}{t}\ln \frac{\left|\right|\mathrm{\δ}{x}_i(x_0,t)\left|\right|}{\left|\right|\mathrm{\δ}{x}_i(x_0,0)\left|\right|}---\left(4\right)$

N maintains system, corresponding n Lyapunov exponential quantity, as long as there is a Lyapunov index to be greater than 0, is in chaos state with regard to illustrative system, two-dimentional Non-Self-Governing Duffing vibrator signal detection system formula (3) represented forms three-dimensional autonomous system, as shown in the formula:

$\left.\begin{array}{c}\stackrel{\·}{x}=v\\ \stackrel{\·}{v}=-\mathrm{kv}+x^3-x^5+F\cos \left(\mathrm{\ωz}\right)+\stackrel{\‾}{s}\left(z\right)\\ \stackrel{\·}{z}=1\end{array}\right\}---\left(5\right)$

Formula (4) is utilized to calculate three Lyapunov index λ corresponding to Duffing vibrator signal detection system respectively ₁, λ ₂and λ ₃;

B) because chaotic motion is very complicated, be difficult to utilize Integral Dimension to go definition, this patent fractal dimension detection technique Main Basis be chaotic motion self have certain can carry out with relatively less degree of freedom the potential order that rationally describes.According to the above-mentioned Lyapunov index that will calculate, obtain a kind of definition of dimension, as follows:

$D=j+\frac{\underset{i=1}{\overset{j}{\mathrm{\Σ}}}{\mathrm{\λ}}_i}{\left|{\mathrm{\λ}}_{j+1}\right|}---\left(6\right)$

Wherein, meet because this definition is not suitable for periodic motion and quasi-periodic motion, then the dimension that delimiting period moves and quasi-periodic motion is corresponding is 2.

With the change of driving force F, Lyapunov index and dimension also change, each motion state of its correspondence, as shown in table 2 below.

Relation between table 2 Duffing vibrator signal detection system and Lyapunov index, dimension

6) the incident wave signal of operating mode one (i.e. intact pipeline) received and the pure noise signal of numerical simulation are input to Duffing vibrator signal detection system respectively, i.e. formula (3), pure noise signal σ e (t) represents, wherein e (t) is a random function analogue noise, σ is noise level, be 0.05 herein, and draw and add the effect diagram that before and after these signals, dimension changes with driving force F, as shown in Figure 3; At F ∈ (0.5,0.52) in scope, no signal inputs and inputs pure noise signal, dimension is 2, expression system is periodic motion or quasi-periodic motion, and the incidence that input receives involves edge echo signal, and dimension is greater than 2, represent that now system is chaos state, and select the driving force F=0.516 of dimension maximum to carry out the identification of ultrasonic guided wave signals.

7) defect tracking utilizing Moving Window function to carry out under different operating mode carries out defect recognition and location

A) Moving Window function is defined as follows:

S ^*＝g(t-nτ)S (7)

$g(t-\mathrm{n\τ})=1,t\∈(\mathrm{n\τ}-\mathrm{\δ},\mathrm{n\τ}+\mathrm{\δ}),n=\frac{\mathrm{\δ}}{\mathrm{\τ}},\frac{\mathrm{\δ}+\mathrm{\τ}}{\mathrm{\τ}},\frac{\mathrm{\δ}+2\mathrm{\τ}}{\mathrm{\τ}}...,\frac{N-\mathrm{\δ}}{\mathrm{\τ}}---\left(8\right)$

Wherein, S represents the full time-domain signal recorded, S ^*represent intercept signal, N represents signal length, and 2 δ represent window length, and τ represents window translational speed.

B) Moving Window function sweep signal

Signal under utilizing Moving Window function to scan four kinds of operating modes, the schematic diagram of its scanning as shown in Figure 4, select 2 δ=50 μ s respectively, τ=5 μ s (can analyze selecting the reason of these two parameters below), namely the dimension of each segment signal is calculated, in order to the order of accuarcy of more sharp defect recognition in this way and location, the time-history curves that the dimension calculated and digital oscilloscope receive (time history curve) is made simultaneously, and make the envelope of dimension, four kinds of operating modes are respectively as shown in Fig. 5 a ~ 5d; As can be seen from Fig. 5 a, operating mode one is intact pipeline, and owing to not having defect, only at incident wave and edge echo place, dimension is greater than 2, and the dimension between incident wave and edge echo equals 2; As can be seen from Fig. 5 b, operating mode two due to defect smaller, time-history curves cannot determine flaw echo, and be greater than 2 by the dimension of the center section calculated, the existence of defect can be determined, as long as owing to there is weak signal, dimension also can be made to become obviously be greater than 2 even if signal is imperfect, the time zone that the dimension therefore calculated is greater than 2 is a scope; As can be seen from Fig. 5 c and Fig. 5 d, operating mode three and operating mode four larger due to defect, time-history curves obviously can determine flaw echo, repeat no more herein.

C) defect location

By the dimension that Fig. 5 a ~ 5d calculates, wider owing to relating to time range, be unfavorable for accurate location; In the present embodiment, two kinds of methods can be selected to carry out accurately determining of waveguide moment, a kind of method selects the envelope peak value place of dimension, as ripple to the moment; Another kind be by the envelope center of dimension as ripple to the moment, wherein defective locations and percentage error, according to following formulae discovery:

$d_x=\frac{t_2-t_1}{t_3-t_1}d---\left(9\right)$

$r=\left|\frac{{2d}_x-d}{2d}\right|\×100\%---\left(10\right)$

Wherein, d _xrepresent that in pipe, defect is from the distance exciting end, d represents duct length, t ₁, t ₂and t ₃represent the moment that incident wave, flaw echo and edge echo receive respectively, r represents relative error.Result of calculation is as shown in table 3 below, can see that to select the envelope peak value of dimension less to the error in moment as ripple from table, therefore all adopts in defect location afterwards and positions in this way.

Defect location under table 3 four kinds of operating modes

D) consider that signal length, 2 δ and τ are on the impact of defect location

When first studying different 2 δ, carry out error analysis that defect location (selecting the envelope peak value place of dimension as ripple to the moment) obtains as shown in Figure 6, when window length is 50 μ s, error r is minimum, and namely damage reason location effect is best.

Suitable in order to select, select different signal lengths, and calculate its dimension, as shown in Figure 7, can find out, when signal length is more than or equal to 10 μ s, the dimension of its correspondence is greater than 2, therefore, window translational speed is only needed to be less than 10 μ s, consider efficiency, τ=5 μ s is selected to carry out the scanning of overall signal, then utilize little speed as τ=1 μ s, provide accurate location, be specially: with operating mode two, three is example, when first utilizing dimension peak value to provide τ=1 μ s, the echo moment in centre position, again according to the velocity of propagation that dispersion curve as shown in Figure 8 calculates, in conjunction with following formula:

2d＝ct (11)

Wherein, c is the velocity of propagation calculated according to dispersion curve, and namely c=5330.58m/s, t represent the moment of the ripple received, and d represents that t pipeline location is from the distance exciting end; Take 2d as horizontal ordinate, dimension is ordinate, makes operating mode two and the operating mode three centre position positioning result under two kinds of window translational speeds (1 μ s and 5 μ s), as illustrated in figures 9 a and 9b; Thus can illustrate, little speed τ=1 μ s is close to defective locations.Therefore, select large velocity sweeping, the pinpoint method of little speed positions ducted little defect.

In sum, the inventive method can carry out effective location to the little defect of ducted Injured level, thus improve the little defect of supersonic guide-wave identification sensitivity, effectively extend sensing range.

The above; be only patent preferred embodiment of the present invention; but protection scope of the present invention is not limited thereto; anyly be familiar with those skilled in the art in scope disclosed in this invention; be equal to according to technical scheme of the present invention and inventive concept thereof and replace or change, all belonged to protection scope of the present invention.

Claims (7)

1., based on the little defect recognition of pipe ultrasonic guided wave and the localization method of dimension, it is characterized in that comprising the following steps:

1) waveform generator generates the signal through the modulation of Hanning window, is amplified, remake the piezoelectric ring for pipeline one side end face by power amplifier, excites the ultrasonic guided wave signals through the modulation of Hanning window, makes supersonic guide-wave travel through all positions of pipeline;

2) receive measured signal by being arranged on piezoelectric patches on pipeline, and by the time-history curves that digital oscilloscope record supersonic guide-wave is propagated in the duct, by digital oscilloscope by the information transmission of record to computing machine;

3) set the incident wave signal of the intact pipeline collected in advance and the pure noise signal of numerical simulation as detection signal, according to centre frequency, the sample frequency of this detection signal, and duffing equation characteristic, structure Duffing vibrator signal detection system;

4) the incident wave signal of the intact pipeline collected in advance and the pure noise signal of numerical simulation are input to Duffing vibrator signal detection system respectively, by compare no signal input, incident wave signal and pure noise signal time, dimension, with the change of driving force amplitude F, determines the F value that can be used for identifying ultrasonic guided wave signals;

5) Moving Window function is defined, select window length 2 δ=50 μ s, window translational speed τ=5 μ s, by Moving Window function scanning measured signal, calculate the dimension of each segment signal, if the dimension between incident wave and edge echo equals 2, then pipeline is intact; If the dimension between incident wave and edge echo is greater than 2, then pipeline defectiveness, enters step 6);

6) in the moment utilizing the envelope peak value determination incident wave of dimension, edge echo and flaw echo to receive, according to the time scale relation between three, defect of pipeline location is carried out;

7) window translational speed τ=1 μ s is utilized accurately to locate ducted little defect.

2. the little defect recognition of pipe ultrasonic guided wave based on dimension according to claim 1 and localization method, is characterized in that: step 1) described through Hanning window modulation ultrasonic guided wave signals expression formula as follows:

$s\left(t\right)=\left[\frac{1}{2}(1-\cos \frac{w_{c}t}{n})\right].\sin \left(w_{c}t\right)---\left(1\right)$

Wherein, n is the single audio frequency number selected, w _c=2 π f _c, f _cfor the centre frequency of measured signal, be 70KHz.

3. the little defect recognition of pipe ultrasonic guided wave based on dimension according to claim 1 and localization method, is characterized in that: step 3) described structure Duffing vibrator signal detection system, specific as follows:

A) choose duffing equation, it comprises the complex state of vibration, fork, chaos, and its expression formula is:

$\stackrel{\·\·}{x}+k\stackrel{\·}{x}-x^3+x^5=F\cos \mathrm{\ωt}---\left(2\right)$

Wherein, k is damping ratio, (-x ³+ x ⁵) be nonlinear resilience item; Fcos ω t is driving force item, and F is driving force amplitude, and ω is driving force angular frequency;

B) set the incident wave signal of the intact pipeline collected in advance and the pure noise signal of numerical simulation as detection signal and its centre frequency is 70KHz, sample frequency is 50M time/second, then sets the ω ≈ 0.439823rad/ μ s of duffing equation, integration step h=0.02 μ s, damping ratio k=0.5, will after input system (1):

$\stackrel{\·\·}{x}+k\stackrel{\·}{x}-x^3+x^5=F\cos \mathrm{\ωt}+\stackrel{\‾}{s}\left(t\right)---\left(3\right)$

Duffing vibrator signal detection system and construction complete.

4. the little defect recognition of pipe ultrasonic guided wave based on dimension according to claim 3 and localization method, is characterized in that: step 5) definition of described dimension is specific as follows:

A) continuous dynamical system x=F (x) is tieed up, in the t=0 moment, with x for n ₀centered by, || δ x (x ₀, 0) || for radius does the sphere of a n dimension; Along with the evolution of time, be namely deformed into the ellipsoid of n dimension at this sphere of t, if half axial length of i-th of this ellipsoid change in coordinate axis direction is, then this system i-th Lyapunov index is:

${\mathrm{\λ}}_i=\underset{t\→\∞}{\lim }\frac{1}{t}\ln \frac{\left|\right|\mathrm{\δ}{x}_i(x_0,t)\left|\right|}{\left|\right|\mathrm{\δ}{x}_i(x_0,0)\left|\right|}---\left(4\right)$

N maintains system, corresponding n Lyapunov exponential quantity, as long as there is a Lyapunov index to be greater than 0, is in chaos state with regard to illustrative system, two-dimentional Non-Self-Governing Duffing vibrator signal detection system formula (3) represented is rewritten as three-dimensional autonomous system, as shown in the formula:

$\left.\begin{array}{c}\stackrel{\·}{x}=v\\ \stackrel{\·}{v}=-\mathrm{kv}+x^3-x^5+F\cos \left(\mathrm{\ωz}\right)+\stackrel{\‾}{s}\left(z\right)\\ \stackrel{\·}{z}=1\end{array}\right\}---\left(5\right)$

Formula (4) is utilized to calculate three Lyapunov index λ corresponding to Duffing vibrator signal detection system respectively ₁, λ ₂and λ ₃;

B) according to the above-mentioned Lyapunov index that will calculate, a kind of definition of dimension is obtained, as follows:

$D=j+\frac{\underset{i=1}{\overset{j}{\mathrm{\Σ}}}{\mathrm{\λ}}_i}{\left|{\mathrm{\λ}}_{j+1}\right|}---\left(6\right)$

Wherein, meet the dimension that delimiting period moves and quasi-periodic motion is corresponding is again 2.

5. the little defect recognition of pipe ultrasonic guided wave based on dimension according to claim 1 and localization method, is characterized in that: step 5) described Moving Window function is defined as follows:

S ^*＝g(t-nτ)S (7)

$g(t-\mathrm{n\τ})=1,t\∈(\mathrm{n\τ}-\mathrm{\δ},\mathrm{n\τ}+\mathrm{\δ}),n=\frac{\mathrm{\δ}}{\mathrm{\τ}},\frac{\mathrm{\δ}+\mathrm{\τ}}{\mathrm{\τ}},\frac{\mathrm{\δ}+2\mathrm{\τ}}{\mathrm{\τ}}...,\frac{N-\mathrm{\δ}}{\mathrm{\τ}}---\left(8\right)$

Wherein, S represents the full time-domain signal recorded, S ^*represent intercept signal, N represents signal length, and 2 δ represent window length, and τ represents window translational speed.

6. the little defect recognition of pipe ultrasonic guided wave based on dimension according to claim 1 and localization method, is characterized in that: step 6) described in carry out defect of pipeline location adopt following formula calculate:

$d_x=\frac{t_2-t_1}{t_3-t_1}d---\left(9\right)$

Wherein, d _xrepresent that in pipe, defect is from the distance exciting end, d represents duct length, t ₁, t ₂and t ₃represent the moment that incident wave, flaw echo and edge echo receive respectively.

7. the little defect recognition of pipe ultrasonic guided wave based on dimension according to claim 1 and localization method, is characterized in that: step 7) the described window translational speed τ=1 μ s that utilizes accurately locates ducted little defect, specific as follows:

When first utilizing dimension peak value to provide window translational speed τ=1 μ s, the echo moment in centre position, then according to the velocity of propagation that dispersion curve calculates, in conjunction with following formula, ducted little defect is accurately located:

2d＝ct (10)

Wherein, c is the velocity of propagation calculated according to dispersion curve, and t represents the moment of the ripple received, and d represents that t pipeline location is from the distance exciting end.