CN101975822B - Damage positioning method for ultrasonic detection of transmission line - Google Patents

Abstract

The invention discloses a damage positioning method for the ultrasonic detection of a transmission line. The time delay of an ultrasonic guided wave signal between a driving point and a damage point is calculated by acquiring a time domain signal of the ultrasonic guided wave detection signal and adopting the technology of wavelet transform to obtain the damage position in the transmission line. The damage positioning method has the advantages of strong generality and accurate positioning, and can be suitable for various ultrasonic guided wave-based detection devices.

Description

A kind of damage positioning method that is used for the power transmission line ultrasound detection

Technical field

The invention belongs to technical field of nondestructive testing, relate in particular to a kind of damage positioning method that is used for the power transmission line ultrasound detection.

Background technology

At industrial circles such as power transmission and transformation, a large amount of various material type power transmission lines that use receive the influence of acid and alkali corrosion environment etc., very easily occur peeling off, slight crack even fracture equivalent damage, cause equipment failure.Ultra-high-tension power transmission line reveals throughout the year and puts in atmosphere through field and city neighborhood, is often bearing the influence of various weather conditions such as wind, ice, temperature.Under sleet or freezing condition, the icing of unsettled power transmission line, high voltage transmission line fixed gantry then can cause huge disaster consequence.Powerline ice-covering and accumulated snow regular meeting cause circuit tripping operation, broken string, fall tower, conductor galloping, insulator arc-over and communicating interrupt etc.The catastrophic failure that causes has thus constituted great threat to the people's life and life security.

According to statistics, 60%-70% line fault is changed by natural conditions and causes that particularly wind, icing and Influence of Temperature are maximum.China's ice damage accident frequently takes place, and increases because of the sleet icing makes the transmission line of electricity load-carrying in many areas, causes broken string, bar accidents such as (towers), has caused enormous economic loss.2004, ice sudden strain of a muscle tripping operation, conductor galloping and the broken string of falling tower accident in a big way appearred in the 500kv circuit of China some areas.At 2008 beginning of the years, China south meets with rare ice and snow disaster, and continuous sleet weather makes that ice covering thickness reaches 50-100mm on a lot of electric power transmission lines, the pylon; Substantially exceeded the technical standard of the net loaded ice covering thickness of general electric power transmission, thereby caused electric tower to collapse, the paralysis of power grid system large tracts of land; And then cause traffic and be obstructed, electric coal is nervous, and communication is not smooth; Power failure is cut off the water, and forest such as is ruined at a series of problems.According to statistics, the direct loss that cause of ice and snow disaster are up to 2530.5 hundred million yuans.Therefore, other damage on-line monitoring technique such as research icing etc., for China's electrical network prevents and controls disaster such as ice and snow, the operational reliability that improves electrical network is significant.Yet traditional transmission line of electricity detection method can not get strong guarantee, and Study on new method is progressively fermented.

Detection kind and detection method for ice covering on transmission lines and other damage defect have developed into tens kinds.Early stage method mainly is that the artificial visually examine checks.The monitoring personnel adopt naked eyes or telescope that lead is observed, and check lead damage or icing.And only just be prone to come to light when perhaps icing is relatively more serious in the lead damage.And because human eye will be done remote observation on quite long lead, easy fatigue, labour intensity is big.Developed aerophotographic method afterwards, and adopted helicopter to fly along transmission line of electricity, and adopted the situation of the observation of artificial visually examine or picture pick-up device and record each abnormity point along the line then, because the restriction of photography, these therefore early stage methods detect very slow, and accuracy is not high.Aspect the online detection Study of New Method of power transmission line, domestic then relatively backward, the Central China University of Science and Technology and Harbin Engineering University have carried out some researchs in this regard, and have obtained certain achievement.

Use elastic wave to detect and not only have the series of advantages that stress wave detects, and detection range is far away, flaw detection efficient is high, and speed is fast.The type that is suitable in addition is also many, can both produce more intense echo to common in the structure like defectives such as layering, folding and crackles.The people such as Saravanos of NASA (NASA) are from theory and experimentally confirmed to utilize elastic wave to detect the delamination damage of compound substance girder construction.Then two kinds of basic elasticity communication mode test material damages of research and utilization of the Percival of Britain's national defence and evaluation studies mechanism and Birt.France ONERA structural damage mechanical system utilizes elasticity to involve the damage The Location that wavelet analysis technology carries out carbon fibre composite.The application of elastic wave technology carries out two independent studies groups that the outstanding research work of damage monitoring is Imperial College at present.The group of Cawley has studied the optimization that directivity Lamb ripple excites, can monitor the damage of metal material in real time, and the group of Soutis then is devoted to the layout and the Signal Processing problem of sensor.At home, many units are also carrying out the research of monitoring structural health conditions.China has at first dropped into the pilot study that certain fund is used to support the healthy OBD problem of large and complex structure system at " 863 " space flight high-tech in the works from the mid-90.Acoustical Inst., Chinese Academy of Sciences answers the photoelastic methods of usefulness such as Chong Fu that the stress distribution of elastic wave has been carried out Direct observation, and they also observe and study elastic wave propagation and scattering.U.S. Simmonds company has developed 2 kinds of independently ultrasonic icing detection measuring systems (IDMS).System adopts pulse---echo technique, can survey freeze phenomenon and ice layer thickness.

From foreign data, it is more that the power transmission line detection technique develops into present method, and like image detection method, infrared detection method, current vortex detection method, robot detection method, wave velocity method etc., but that all is that all right is ripe for great majority, all is in the starting stage.For icing and other damage monitoring of long distance transmission line, classic method can not meet the demands because its maximum detects the restriction of distance and accuracy of detection and engineering adaptability.And based on the damage detecting method of elastic wave, for the solution of this problem provides strong technical support.

When elastic wave was propagated in structure, the various damages of structure can cause that stress is concentrated, crack propagation, and these and damage border on every side all can cause the scattering of the elastic wave signal of propagating in the structure and the absorption of energy.Just be based on this phenomenon, elastic wave can be used for the damage in the structure is monitored.Elastic wave monitoring method initiatively can be divided into four processes: (1) excites the elastic wave signal with specific signal excitation driving element in structure.(2) adopt piezoelectric sensor or other sensors to accept the elastic wave signal at the diverse location of structure.(3) the elastic wave signal that receives is carried out analyzing and processing, extract the damage characteristic parameter.(4) adopt advanced location algorithm that damage is positioned.The characteristics of this technology are to detect distance, and efficient is high.At present, based on the damage detection system of elastic wave in being similar to the one-dimentional structure of power transmission line (fields such as petroleum pipe line detection, rail damage check) obtain application of practical project.Therefore the elastic wave detection method is applied in powerline ice-covering and other damage check and has important researching value and potentiality.

Summary of the invention

The object of the invention provides a kind of damage positioning method that is used for the power transmission line ultrasound detection exactly for addressing the above problem, and it calculates the time delay of ultrasonic guided wave detection signals based on wavelet transformation technique, thereby realizes power transmission line damage positioned detection.

For realizing above-mentioned purpose, the present invention adopts following technical scheme:

A kind of damage positioning method that is used for the power transmission line ultrasound detection; It is provided with at least two drivers and encourages and receive signal on transmission line of electricity; Gather excitation and transducing signal under transmission line of electricity health status and the faulted condition respectively; Thereby obtain transducing signal difference signal under the two states, this difference signal is the damage reflected signal; The characteristics of utilizing velocity of wave after reflecting, to remain unchanged basically through damage, utilize formula:

$\frac{L_0}{t}=\frac{L_0+2x}{t+\mathrm{\Δt}}$

Wherein, L ₀Be the distance of excitation driver to the reception driver, x is damaged to the distance that receives the driver mid point, and t is that signal passes to the time that receives driver from excitation driver, and Δ t is delay time of arrival of damage reflected signal and health signal, L ₀Be definite value, t, Δ t obtain through damage and health signal are carried out the Gabor wavelet transformation respectively; The time delay Δ t of two column signals is pairing time of peak-peak poor of the two Gabor wavelet transformation, can damage the location.

Said Gabor wavelet transformation is following:

The Gabor wavelet function is defined as ${\mathrm{\&psi;}}_m\left(t\right)=\frac{1}{\sqrt[4]{{\mathrm{\&sigma;}}^2\mathrm{\&pi;}}}\mathrm{Exp}(-\frac{{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="HDA0000031926920000012.png,HDA0000031926920000013.png"\; state="\{\{state\}\}">t</figure-callout>}}^2}{{2\mathrm{\&sigma;}}^2}+j{\mathrm{\&omega;}}_{c}t)$

Wherein, σ, ω _cBe positive constant, t is that signal passes to the time that receives driver from excitation driver, and its Fourier transform is:

${\widehat{\mathrm{\&psi;}}}_m\left(\mathrm{\&omega;}\right)=\frac{\sqrt{2\mathrm{\&pi;}}}{\sqrt[4]{\mathrm{\&pi;}{\mathrm{\&sigma;}}^2}}\exp (-\frac{1}{2{\mathrm{\&sigma;}}^2{(\mathrm{\&omega;}-{\mathrm{\&omega;}}_c)}^2})$

Regard the Gabor wavelet function as the center at the Gaussian of t=0 window function; The center of its Fourier transform is ω=ω _c, and function The center be t=b, wherein a is a scale factor, b is a shift factor, its Fourier transform

The center be ω=ω _c/ a, then Gabor wavelet transformation CWT _f(a, b) representative function f (t) is at t=b, ω=ω _cNear/the a time-frequency composition.

Consideration is propagated along the x direction, the divergent wave that same units amplitude, two different harmonic waves of angular frequency constitute, wherein, k ₁, k ₂Be respectively the wave number of these two harmonic waves, w ₁And w ₂Be respectively the angular frequency of these two harmonic waves:

$u(x,t)=e^{-i(k_{1}x-{\mathrm{\&omega;}}_{1}t)}+e^{-i(k_{2}x-{\mathrm{\&omega;}}_{2}t)}$

If

$\left.\begin{array}{c}({\mathrm{<figure-callout\; id="1"\; label="k"\; filenames="HDA0000031926920000012.png,HDA0000031926920000014.png"\; state="\{\{state\}\}">k</figure-callout>}}_1+k_2)/2={\mathrm{<figure-callout\; id="0"\; label="k"\; filenames="HDA0000031926920000012.png,HDA0000031926920000013.png"\; state="\{\{state\}\}">k</figure-callout>}}_0;(k_1-k_2)/2=\mathrm{\&Delta;k};\\ ({\mathrm{\&omega;}}_1+{\mathrm{\&omega;}}_2)/2={\mathrm{\&omega;}}_0;({\mathrm{\&omega;}}_1-{\mathrm{\&omega;}}_2)/2=\mathrm{\&Delta;\&omega;};\end{array}\right\}$

Then following formula can be expressed as:

$u(x,t)=2\cos (\mathrm{\&Delta;kx}-\mathrm{\&Delta;\&omega;t})e^{-i(k_0-{\mathrm{\&omega;}}_{0}t)}$

Can find out that from following formula the stress wave of final generation is made up of two parts: the carrier wave of partly being represented by exponential function is with phase velocity c _p=ω ₀/ k ₀Propagate; Another part modulating wave then is made up of cosine function, with group velocity c _g=Δ ω/Δ k propagates.

Use the Gabor small echo to u (x, t) do wavelet transformation and get:

$\mathrm{CW}{T}_u(x,a,b)=\sqrt{a}\{e^{-i(k_{1}x-{\mathrm{\&omega;}}_{1}b)}{{\widehat{\mathrm{\&psi;}}}_m}^{*}\left(a{\mathrm{\&omega;}}_1\right)+e^{-i(k_{2}x-{\mathrm{\&omega;}}_{2}b)}{{\widehat{\mathrm{\&psi;}}}_m}^{*}\left(a{\mathrm{\&omega;}}_2\right)\}$

Amplitude behind the wavelet transformation is:

$\left|\mathrm{CW}{T}_u(x,a,b)\right|=\sqrt{a}\{{\left[{\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_1\right)\right]}^2+{\left[{\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_2\right)\right]}^2+2{\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_1\right){\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_2\right)\cos (2\mathrm{\&Delta;kx}-2\mathrm{\&Delta;\&omega;b}){\}}^{1/2}$

When enough hour of Δ ω (Δ ω is not a definite number, that is, if formula thereafter is false, explain that the value of Δ ω is still inadequately little, continue value, set up up to the formula of back, explain that then it is enough little), satisfied:

${\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_1\right)\&ap;{\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_2\right)\&ap;{\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_0\right)$

Obtain Gabor wavelet transformation (CWT _U) mould:

$\left|\left(\mathrm{CW}{T}_u\right)(x,a,b)\right|\&ap;\sqrt{\mathrm{ab}}\left|{\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_c\right)\right|{[1+\cos (2\mathrm{\&Delta;kx}-2\mathrm{\&Delta;\&omega;b})]}^{1/2}$

Mould behind wavelet transformation is at a=ω _c/ ω ₀, b=(Δ k/ Δ ω) x=x/c _gThe time obtain maximal value, promptly (a, b) peak value on the plane corresponding frequency be ω ₀=ω _cRipple bag during/a is with group velocity c _gThe due in of propagating, the time delay Δ t of two column signals are pairing time of peak-peak poor of the two Gabor wavelet transformation.

Said driver is a piezoelectric patches.

The present invention adopts driver to excite the active monitoring signal at body structure surface, and meanwhile sensor receives structural response signals in other places on same surface, and signal is analyzed, and in view of the above damage in the structure is monitored, and principle is as shown in Figure 1.When damage appears in structure, can cause the scattering of the monitor signal of in structure, propagating and the absorption of energy.Damage signal and health signal through monitoring compare, and then carry out the structural damage location.This paper adopts two to adopt piezoelectric patches (PZT) to encourage and receive signal respectively; Gather excitation and transducing signal under structural health conditions and the faulted condition respectively; Can obtain the difference signal of transducing signal under the two states, this difference signal is the damage reflected signal.Suppose that velocity of wave remains unchanged basically, then can obtain following relational expression after through the damage reflection:

$\frac{L_0}{t}=\frac{L_0+2x}{t+\mathrm{\&Delta;t}}$

L wherein ₀Be the distance of excitation PZT to sensing PZT, x is damaged to the distance that receives the PZT mid point, and t is the time of signal from encouraging PZT to pass to sensing PZT, and Δ t is for postponing the time of arrival of damage reflected signal and health signal.Because L ₀Necessarily, t, Δ t can obtain through monitor signal is carried out the Gabor wavelet transformation.

The continuous wavelet transform of function f (t) defines as follows:

$\mathrm{CWT}(a,b)={\&Integral;}_{-\&infin;}^{\&infin;}f\left(t\right){\mathrm{\&psi;}}_{a,b}^{*}\left(t\right)\mathrm{dt}$

In the formula; is wavelet function, and symbol " * " is represented its conjugater transformation.Wavelet function is obtained through after flexible, the translation by female small echo Ψ (t), and wherein a is called scale factor, and b is called shift factor.Under enabled condition, through a, b moves, and just can obtain the T/F window of a flexibility and changeability, makes when high " centre frequency ", and window narrows down automatically, and when low " centre frequency ", broadens automatically, thereby reaches the effect of time frequency analysis.

The wavelet function that this paper selects for use is the Gabor small echo, and it provides better time frequency resolution.The Gabor wavelet function defines as follows:

${\mathrm{\&psi;}}_m\left(t\right)=\frac{1}{\sqrt[4]{{\mathrm{\&sigma;}}^2\mathrm{\&pi;}}}\exp (-\frac{{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="HDA0000031926920000012.png,HDA0000031926920000013.png"\; state="\{\{state\}\}">t</figure-callout>}}^2}{2{\mathrm{\&sigma;}}^2}+{\mathrm{j\&omega;}}_{c}t)$

Its Fourier transform is:

${\widehat{\mathrm{\&psi;}}}_m\left(\mathrm{\&omega;}\right)=\frac{\sqrt{2\mathrm{\&pi;}}}{\sqrt[4]{\mathrm{\&pi;}{\mathrm{\&sigma;}}^2}}\exp \left(\frac{1}{{2\mathrm{\&sigma;}}^2{(\mathrm{\&omega;}-{\mathrm{\&omega;}}_c)}^2}\right)$

σ, ω _cBe positive constant.The Gabor wavelet function can be regarded as the center at the Gaussian of t=0 window function; The center of its Fourier transform is ω=ω _cAnd function

The center be t=b, wherein a is a scale factor, b is a shift factor, its Fourier transform

The center be ω=ω _c/ a.Gabor wavelet transformation CWT _f(a, b) representative function f (t) is at t=b, ω=ω _cNear/the a time-frequency composition.Make ω in the experiment _c=2 π, then 1/a equals centre frequency f=ω/2 π.

Consideration is propagated along the x direction, the divergent wave that same units amplitude, two different harmonic waves of angular frequency constitute:

$u(x,t)=e^{-i(k_{1}x-{\mathrm{\&omega;}}_{1}t)}+e^{-i(k_{2}x-{\mathrm{\&omega;}}_{2}t)}$

Use the Gabor small echo to u (x, t) do wavelet transformation and get:

${\mathrm{CWT}}_u(x,a,b)=\sqrt{a}\{e^{-i(k_{1}x-{\mathrm{\&omega;}}_{1}b)}{{\widehat{\mathrm{\&psi;}}}_m}^{*}\left(a{\mathrm{\&omega;}}_1\right)+e^{-i(k_{2}x-{\mathrm{\&omega;}}_{2}b)}{{\widehat{\mathrm{\&psi;}}}_m}^{*}\left(a{\mathrm{\&omega;}}_2\right)\}$

If Δ ω enough little (Δ ω is not a definite number, that is if thereafter formula is false, explain that the value of Δ ω is still inadequately little, continue value, the formula establishment up to the back explains that then it is enough little) obtains Gabor wavelet transformation (CWT _U) mould:

$\left|\left({\mathrm{CWT}}_u\right)(x,a,b)\right|\&ap;\sqrt{\mathrm{ab}}\left|{\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_c\right)\right|{[1+\cos (2\mathrm{\&Delta;kx}-2\mathrm{\&Delta;\&omega;b})]}^{1/2}$

This result shows: the mould behind the wavelet transformation is at a=ω _c/ ω ₀, b=(Δ k/ Δ ω) x=x/c _gThe time obtain maximal value, promptly (a, b) peak value on the plane corresponding frequency be ω ₀=ω _cRipple bag during/a is with group velocity c _gThe due in of propagating.

According to above conclusion, the time delay Δ t of two column signals is pairing time of peak-peak poor of the two Gabor wavelet transformation.The time delay of utilizing the difference in this moment to make is herein just damaged the location.

The invention has the beneficial effects as follows:, can realize the propagation of elastic wave single-mode, and can be applied in the monitoring structural health conditions through regulating excitation frequency.It is better to adopt the Gabor wavelet transformation to measure damage reflected signal carryover effects time of arrival, and the damage bearing accuracy is higher.

Description of drawings

Fig. 1 damages the synoptic diagram of location for the present invention;

Fig. 2 a is the time domain and the frequency domain figure of pumping signal;

Fig. 2 b is the time domain and the frequency domain figure of pumping signal;

Fig. 3 a is monitor signal figure;

Fig. 3 b is monitor signal figure;

Fig. 3 c is monitor signal figure;

Fig. 3 d is the corresponding Gabor wavelet transformation figure of Fig. 3 a;

Fig. 3 e is the corresponding Gabor wavelet transformation figure of Fig. 3 b;

Fig. 3 f is the corresponding Gabor wavelet transformation figure of Fig. 3 c.

Wherein, 1 transmission line of electricity, 2 piezoelectric patches, 3 masses.

Embodiment

Below in conjunction with accompanying drawing and embodiment the present invention is further specified.

Among Fig. 1; It is provided with at least two piezoelectric patches 2 and encourages and received signal on transmission line of electricity 1; Gather excitation and transducing signal under transmission line of electricity health status and the faulted condition respectively, thereby obtain transducing signal difference signal under the two states, this difference signal is the damage reflected signal; The characteristics of utilizing velocity of wave after reflecting, to remain unchanged basically through damage, utilize formula:

$\frac{L_0}{t}=\frac{{\mathrm{<figure-callout\; id="0"\; label="L"\; filenames="HDA0000031926920000012.png,HDA0000031926920000013.png"\; state="\{\{state\}\}">L</figure-callout>}}_0+2x}{t+\mathrm{\&Delta;t}}$

L wherein ₀Be the distance of excitation piezoelectric patches to the reception piezoelectric patches, x is damaged to the distance that receives the piezoelectric patches mid point, and t is that signal passes to the time that receives piezoelectric patches from the excitation piezoelectric patches, and Δ t is for postponing L the time of arrival of damaging reflected signal and health signal ₀Be definite value, t, Δ t obtain through damage and health signal are carried out the Gabor wavelet transformation respectively; The time delay Δ t of two column signals is pairing time of peak-peak poor of the two Gabor wavelet transformation, can damage the location.

Said Gabor wavelet transformation is following:

The Gabor wavelet function is defined as ${\mathrm{\&psi;}}_m\left(t\right)=\frac{1}{\sqrt[4]{{\mathrm{\&sigma;}}^2\mathrm{\&pi;}}}\mathrm{Exp}(-\frac{{\mathrm{<figure-callout\; id="2"\; label="t"\; filenames="HDA0000031926920000012.png,HDA0000031926920000013.png"\; state="\{\{state\}\}">t</figure-callout>}}^2}{{2\mathrm{\&sigma;}}^2}+j{\mathrm{\&omega;}}_{c}t)$

Wherein, σ, ω _cBe positive constant, t is that signal passes to the time that receives driver from excitation driver; Its Fourier transform is:

${\widehat{\mathrm{\&psi;}}}_m\left(\mathrm{\&omega;}\right)=\frac{\sqrt{2\mathrm{\&pi;}}}{\sqrt[4]{{\mathrm{\&pi;\&sigma;}}^2}}\exp (-\frac{1}{{2\mathrm{\&sigma;}}^2{(\mathrm{\&omega;}-{\mathrm{\&omega;}}_c)}^2})$

Regard the Gabor wavelet function as the center at the Gaussian of t=0 window function; The center of its Fourier transform is ω=ω _c, and function

The center be t=b, wherein a is a scale factor, b is a shift factor, its Fourier transform

The center be ω=ω _c/ a, then Gabor wavelet transformation CWT _f(a, b) representative function f (t) is at t=b, ω=ω _cNear/the a time-frequency composition.

Consideration is propagated along the x direction, the divergent wave that same units amplitude, two different harmonic waves of angular frequency constitute, wherein, k ₁, k ₂Be respectively the wave number of these two harmonic waves, w ₁And w ₂Be respectively the angular frequency of these two harmonic waves:

$u(x,t)=e^{-i(k_{1}x-{\mathrm{\&omega;}}_{1}t)}+e^{-i(k_{2}x-{\mathrm{\&omega;}}_{2}t)}$

Use the Gabor small echo to u (x, t) do wavelet transformation and get:

${\mathrm{CWT}}_u(x,a,b)=\sqrt{a}\{e^{-i(k_{1}x-{\mathrm{\&omega;}}_{1}b)}{{\widehat{\mathrm{\&psi;}}}_m}^{*}\left(a{\mathrm{\&omega;}}_1\right)+e^{-i(k_{2}x-{\mathrm{\&omega;}}_{2}b)}{{\widehat{\mathrm{\&psi;}}}_m}^{*}\left(a{\mathrm{\&omega;}}_2\right)\}$

When enough hour of Δ ω (Δ ω is not a definite number, that is if formula thereafter is false, explain that the value of Δ ω is still inadequately little, continue value, the formula establishment up to the back explains that then it is enough little), obtain Gabor wavelet transformation (CWT _U) mould:

$\left|\left({\mathrm{CWT}}_u\right)(x,a,b)\right|\&ap;\sqrt{\mathrm{ab}}\left|{\widehat{\mathrm{\&psi;}}}_m\left(a{\mathrm{\&omega;}}_c\right)\right|{[1+\cos (2\mathrm{\&Delta;kx}-2\mathrm{\&Delta;\&omega;b})]}^{1/2}$

Mould behind wavelet transformation is at a=ω _c/ ω ₀, b=(Δ k/ Δ ω) x=x/c _gThe time obtain maximal value, promptly (a, b) peak value on the plane corresponding frequency be ω ₀=ω _cRipple bag during/a is with group velocity c _gThe due in of propagating, the time delay Δ t of two column signals are pairing time of peak-peak poor of the two Gabor wavelet transformation.

Embodiment:

Adopt two arc piezoelectric patches 2 to encourage and receive elastic wave respectively, this representative engineering problem is located in the damage of one-dimentional structure carried out experimental study.Whole experimental system comprises power amplifier, function signal generator, computing machine, stimulus sensor, receiving sensor, and mass.

Be pasted with a mass 3 on the extended line as excitation and two piezoelectric patches of sensor respectively, damaging with the icing of mimic transmission line.Excitation piezoelectric patches internal diameter is 22.3mm, thickness 3.2mm, long 30mm.What pumping signal adopted is seven crest Sine Modulated narrow band signals, and peak value is 10V, shown in Fig. 2 a, Fig. 2 b.Experimental studies results through analyzing the different frequency range scope finds that the communication mode of 30KHZ elastic wave in power transmission line is fairly simple, and relatively more responsive to microlesion, it is 30kHz that this paper adopts the pumping signal centre frequency.Gather excitation and transducing signal under structural health conditions (not placing mass 3) and the faulted condition (placing mass 3) in the experiment respectively, can obtain the difference signal of transducing signal under the two states, this difference signal is the damage reflected signal.The time delay that obtains under the faulted condition through the Gabor wavelet transformation then positions.

Fig. 3 a-Fig. 3 f is the experimental result in power transmission line.The left side is the experimental monitoring original signal among Fig. 3 a-Fig. 3 c, and the right side is its corresponding Gabor wavelet transform signal.Fig. 3 a is the structural response signal (health signal) before mass applies; Fig. 3 b is the structural response signal (damage signal) that adds behind the mass; Fig. 3 c is that health signal subtracts each other the difference signal that is obtained with damage signal, and Fig. 3 d, Fig. 3 e, Fig. 3 f are respectively its corresponding Gabor wavelet transformation.Can clearly be seen that from Fig. 3 a transducing signal is with A ₀Pattern is main, S before ₀Mode signal has received inhibiting effect.Comparison diagram 3c and Fig. 3 a, Fig. 3 b two figure can clearly see the damage reflected signal that postpones arrival.In order to improve locating accuracy, utilize the Gabor wavelet transformation to calculate the ripple due in, utilize formula (7) can calculate damage position again.

Table 1 has shown the data result of four repeated experiments, and the actual damage center is x=30cm, and is visible, and the damage bearing accuracy is good, and experimental repeatability is fine.Error maybe elastic wave signal propagation attenuation causes in structure and the signal reflex zone does not cause on same straight line.

Table 1 damage positioning experiment result

Can know through above research,, can realize the propagation of elastic wave single-mode, and can be applied in the monitoring structural health conditions through regulating excitation frequency.It is better to adopt the Gabor wavelet transformation to measure damage reflected signal carryover effects time of arrival, and the damage bearing accuracy is higher.

In the end excitation elastic wave of power transmission line, because elastic wave is very little along the travel path decay, so it can propagate distance very far away along power transmission line.If receiving transducer be positioned at power transmission line certain a bit, then receive the globality information that signal has just comprised transmission line of electricity.Therefore, this technology is compared with other traditional method and had following advantage: propagation distance is far away, detection efficiency is high, the simple easy operating of equipment, and security is high, is very suitable for ice covering on transmission lines and other damage check.

Claims (3)

1. damage positioning method that is used for the power transmission line ultrasound detection; It is characterized in that; It is provided with at least two drivers and encourages and receive signal on transmission line of electricity; Gather excitation and transducing signal under transmission line of electricity health status and the faulted condition respectively, thereby obtain transducing signal difference signal under the two states, this difference signal is the damage reflected signal; The characteristics of utilizing velocity of wave after reflecting, to remain unchanged basically through damage, utilize formula:

$\frac{L_0}{t}=\frac{L_0+2x}{t+\mathrm{\&Delta;t}}$

L wherein ₀Be the distance of excitation driver to the reception driver, x is damaged to the distance that receives the driver mid point, and t is that signal passes to the time that receives driver from excitation driver, and Δ t is delay time of arrival of damage reflected signal and health signal, L ₀Be definite value, t, Δ t obtain through damage and health signal are carried out the Gabor wavelet transformation respectively; The time delay Δ t of two column signals is pairing time of peak-peak poor of the two Gabor wavelet transformation, damages the location.

2. the damage positioning method that is used for the power transmission line ultrasound detection as claimed in claim 1 is characterized in that, said Gabor wavelet transformation is following:

The Gabor wavelet function is defined as ${\mathrm{\&psi;}}_m\left(t\right)=\frac{1}{\sqrt[4]{{\mathrm{\&sigma;}}^2\mathrm{\&pi;}}}\mathrm{Exp}(-\frac{t^2}{2{\mathrm{\&sigma;}}^2}+j{\mathrm{\&omega;}}_{c}t)$

Wherein, σ, ω _cBe positive constant, t is that signal passes to the time that receives driver from excitation driver, and its Fourier transform is:

${\widehat{\mathrm{\&psi;}}}_m\left(\mathrm{\&omega;}\right)=\frac{\sqrt{2\mathrm{\&pi;}}}{\sqrt[4]{{\mathrm{\&pi;\&sigma;}}^2}}\exp (-\frac{1}{2{\mathrm{\&sigma;}}^2{(\mathrm{\&omega;}-{\mathrm{\&omega;}}_c)}^2})$

Regard the Gabor wavelet function as the center at the Gaussian of t=0 window function; The center of its Fourier transform is ω=ω _c, and function

The center be t=b, its Fourier transform

The center be ω=ω _c/ a, then Gabor wavelet transformation CWT _f(a, b) representative function f (t) is at t=b, ω=ω _cNear/the a time-frequency composition;

Consideration is propagated along the x direction, the divergent wave that same units amplitude, two different harmonic waves of angular frequency constitute, wherein, k ₁, k ₂Be respectively the wave number of these two harmonic waves, w ₁And w ₂Be respectively the angular frequency of these two harmonic waves:

$u(x,t)=e^{-i(k_{1}x-{\mathrm{\&omega;}}_{1}t)}+e^{-i(k_{2}x-{\mathrm{\&omega;}}_{2}t)}$

Use the Gabor small echo to u (x, t) do wavelet transformation and get:

${\mathrm{CWT}}_u(x,a,b)=\sqrt{a}\{e^{-i(k_{1}x-{\mathrm{\&omega;}}_{1}b)}{\widehat{\mathrm{\&psi;}}}_m*\left({\mathrm{a\&omega;}}_1\right)+e^{-i(k_{2}x-{\mathrm{\&omega;}}_{2}b)}{\widehat{\mathrm{\&psi;}}}_m*\left({\mathrm{a\&omega;}}_2\right)\}$

As enough hour of Δ ω, obtain Gabor wavelet transformation CWT _UMould:

$\left|\left({\mathrm{CWT}}_u\right)(x,a,b)\right|\&ap;\sqrt{\mathrm{ab}}\left|{\widehat{\mathrm{\&psi;}}}_m\left({\mathrm{a\&omega;}}_c\right)\right|{[1+\cos (2\mathrm{\&Delta;kx}-2\mathrm{\&Delta;\&omega;b})]}^{1/2}$

Mould behind wavelet transformation is at a=ω _c/ ω ₀, b=(Δ k/ Δ ω) x=x/c _gThe time obtain maximal value, promptly (a, b) peak value on the plane corresponding frequency be ω ₀=ω _cRipple bag during/a is with group velocity c _gThe due in of propagating, the time delay Δ t of two column signals are pairing time of peak-peak poor of the two Gabor wavelet transformation.

3. the damage positioning method that is used for the power transmission line ultrasound detection as claimed in claim 1 is characterized in that, said driver is a piezoelectric patches.

Images