CN103278851B - State testing method buried by a kind of submarine pipeline - Google Patents

Abstract

The invention discloses a kind of submarine pipeline and bury state testing method, described method comprises: utilize the grouping voice data after Fast Fourier Transform (FFT) to make the spectrogram of subset of array; The statistic histogram of spectrum distribution is made by spectrogram; The comparison interval of setting grouping voice data, by subscript difference for histogram vectors that the grouping voice data of Δ l is corresponding subtract each other and ask its mould square, obtain difference data sequence; Obtain fitting result function by described difference data sequence and burying pipeline transition section voice data sampling number, and carry out peakvalue's checking; The statistic histogram corresponding with peak value is judged, so identify pipeline bury situation.This method shortens sense cycle, reduce detection difficulty and cost, and provide pipeline comparatively accurately and bury the criterion with naked leakage state completely, workmen can find the naked leakage of pipeline timely, avoid loss economically, meet the multiple needs in practical application.

Description

Submarine pipeline burying condition detection method

Technical Field

The invention relates to the field of pipeline detection, in particular to a submarine pipeline burying condition detection method.

Background

With the expansion of the marine oil exploitation scale in China, the construction of seabed oil and gas pipelines gradually enters a peak, and the seabed pipelines of medium sea oil, medium petrochemical industry and other large companies only in Bohai Bay reach hundreds of kilometers. The submarine pipeline is a marine petroleum project with high investment and high risk, and plays an important role in the production of offshore oil and gas fields. The marine environment in which the submarine pipeline is located is very complex, and the submarine pipeline is influenced by natural factors such as storms, ocean currents and storm tides, so that potential safety hazards such as bare and suspended exist generally. The submarine pipeline is exposed in a hanging mode for a long time, and is easy to generate fatigue fracture when being impacted by wind waves and sea currents; meanwhile, due to the damage of artificial factors such as offshore engineering construction, fishing operation and ship anchoring, submarine pipeline damage accidents sometimes occur. Once the submarine oil and gas pipeline leaks, the offshore oil and gas field stops production, the marine environment is polluted, ecological disasters are caused, the first-aid repair cost is high, even the offshore oil platform explodes, and huge economic losses are brought to enterprises and countries. Therefore, scientific detection of the state of the submarine pipeline is urgent and is important for safe operation of offshore oil.

At present, the main acoustic methods for detecting the submarine pipeline commonly used at home and abroad comprise: the system comprises a multi-beam depth sounder, a side-scan sonar, a shallow stratum profiler, an underwater camera, an ocean magnetometer and the like, and is used for completing the investigation of pipeline burying conditions under water and a seabed mud surface and submarine geological conditions of a routing area. The multi-beam depth sounder and the side-scan sonar achieve the purpose of investigating the exposure position and the suspended height of the suspended submarine pipeline by detecting the water depth and the submarine topography; the shallow stratum profiler can obtain high-frequency and low-frequency shallow stratum profile data, survey of the buried depth of the buried pipeline and the type and thickness of the overlying sediments is realized, multi-beam sounding cannot display the state of the completely buried pipeline, and the thickness of the overlying sediments needs to be determined by matching with the shallow stratum profiler. The submarine pipeline detected by the underwater camera has strong intuition, but the visibility is very low. Marine magnetometers may be used to detect the presence of a pipeline, but may not detect the spatial state of the pipeline.

In the process of implementing the invention, the inventor finds that the prior art has at least the following disadvantages and shortcomings:

the methods have respective defects, need comprehensive utilization and combined analysis, and the detection difficulty rises sharply along with the increase of the depth of the submarine pipeline. This leads to the problems of complex detection task, high difficulty, high cost and long period.

Disclosure of Invention

The invention provides a submarine pipeline burying condition detection method, which shortens the detection period, reduces the detection difficulty and cost, and is described in detail in the following description:

a method of subsea pipeline burial condition detection, the method comprising:

using packet voice data X after fast Fourier transform_l(k) Making a spectrogram of the array subset; making a statistical histogram of the frequency spectrum distribution through a spectrogram;

setting a comparison interval of the grouped sound data, subtracting histogram vectors corresponding to the grouped sound data with subscript difference delta l and solving the square of a modulus of the histogram vectors to obtain a difference data sequence;

acquiring a fitting result function through the difference data sequence and the number of sampling points of the sound data of the pipeline burying transition section, and performing peak value detection;

and judging the statistical histogram corresponding to the peak value, and further identifying the burying condition of the pipeline.

Grouping sound data X after the fast Fourier transform_l(k) Before the step of making the spectrogram of the array subset, the method further comprises:

fixing a sound sensor at any position in a detector in the pipeline to acquire sound data in the pipeline and transmitting the sound data to an upper computer; fast Fourier Transform (FFT) is performed on each packet of voice data to obtain X_l(k)。

The step of making the statistical histogram of the spectral distribution by the spectrogram specifically comprises the following steps:

will [0, f/2 ] of the spectrogram]The interval is divided into M sections; for an abscissa value f within each segment interval_cCalculating the sum of the corresponding ordinate values; and acquiring a statistical histogram vector.

The step of obtaining the fitting result function through the difference data sequence and the number of sampling points of the sound data of the pipeline burying transition section specifically comprises the following steps:

retrieving a maximum value in the difference data sequence, assigning the maximum value to G_maxSetting a threshold coefficient, if there is continuous data greater than a threshold G_maxThen the sequential data is organized into an array subset.

In addition, the step of judging the statistical histogram corresponding to the peak value and further identifying the burying condition of the pipeline specifically comprises the following steps:

1) determiningCorresponding statistical histogram vector H₁For discrete data in the vector { h }₁,h₂,…,h_MPerforming least square fitting of a quadratic polynomial, detecting a peak value, and recording an abscissa value l' at the peak value;

2) when a peak is detected and l' < gammam, gamma is a low frequency threshold coefficient indicating that the pipeline is entering a fully buried state;

3) when two peaks are detected and l₁’<γM、l₂’>M is high frequency threshold coefficient, and is more than gamma, which indicates that the pipeline enters an exposed state.

The technical scheme provided by the invention has the beneficial effects that: this method relies on an in-pipe inspection system for in-pipe inspection, as opposed to other out-of-pipe inspection methods. The internal detection system has the advantages of weight of only several kilograms to dozens of kilograms, low manufacturing cost, small size and short detection period of only several hours; the inner detection system freely advances in the pipeline under the pushing of the conveying medium, does not influence the normal transportation of the pipeline, and is irrelevant to the depth of the submarine pipeline and the external marine environment; the method shortens the detection period, reduces the detection difficulty and cost, accurately provides the judgment standard of the complete burying and bare leakage states of the pipeline, and can be used for timely finding the bare leakage of the pipeline by constructors, thereby avoiding economic loss and meeting various requirements in practical application.

Drawings

FIG. 1 is a graph of a spectrum of an acoustic signal;

FIG. 2 is a diagram illustrating a sound spectrum concentrated in a low frequency region;

FIG. 3 is a schematic diagram of the distribution of the sound spectrum in both the low frequency and high frequency regions;

fig. 4 is a flowchart of a method for detecting a buried condition of a submarine pipeline.

Detailed Description

In order to make the objects, technical solutions and advantages of the present invention more apparent, embodiments of the present invention will be described in detail with reference to the accompanying drawings.

101: fixing a sound sensor at any position in a detector in the pipeline to acquire sound data in the pipeline and transmitting the sound data to an upper computer;

the detailed operation of the step is as follows: fixing the sound sensor at any position in the detector in the pipeline, putting the inner detector into the pipeline for inspection, measuring sound data in the pipeline, downloading the sound data recorded by the inner detector to an upper computer after inspection, and performing data processing.

The embodiment of the present invention does not limit the types of the sound sensor and the detector in the pipeline, and any device capable of performing the above functions may be used, for example: the in-pipe detector may be a cylindrical or spherical inner detector or the like.

102: the upper computer divides the obtained sound data into groups to obtain grouped sound data { D_j}_l；

The detailed operation of the step is as follows: sound obtained by an upper computerThe sound data is { D_iI =1,2,3 … N, N being the total number of points of the sound data. With N₃For step length, grouping the voice data, the number of points is less than N₃Is truncated to obtain the grouped sound data { D }_j|j=0,1,2…N₃-1}_l，l=1,2,...,N\N₃Wherein the symbol "\" represents a whole division.

$N_1=\frac{L}{v}f$

N₃=αN₁

Wherein L is the length of the pipeline burying transition section and can be obtained from construction data, v is the average detection speed of the internal detector, f is the sampling frequency of the sound sensor, and N is the sampling frequency of the sound sensor₁The number of sampling points of the sound data of the transition section is buried in the pipeline, α is 0.2, the number can be properly changed according to the actual situation, but the number is not suitable to be too large, for example, N is 100₃When the number is 6, the obtained packet sound data is:

{D₀,D₁,D₂,D₃,D₄,D₅}₁,{D₀,D₁,D₂,D₃,D₄,D₅}₂,…,{D₀,D₁,D₂,D₃,D₄,D₅}₁₆。

103: fast Fourier Transform (FFT) is performed on each packet of voice data to obtain X_l(k)。

$X_l\left(k\right)=\mathrm{DFT}\left[{\left\{D_j\right\}}_l\right]=\underset{j=0}{\overset{N_3-1}{\mathrm{\&Sigma;}}}{D}_j{\left(W_{N_3}\right)}^{\mathrm{kj}},0\&le;k\&le;N_3-1$

$W_{N_3}=e^{-j(2\mathrm{\&pi;}/N_3)}$

Since Discrete Fourier Transform (DFT) and Fast Fourier Transform (FFT) are only different in operation speed and the operation results are the same, the DFT formula is used instead of the FFT to express the transformed results. The index l indicates that the ith subset is processed.

104: using fast Fourier transform X_l(k) Making a spectrogram of the array subset;

the method comprises the following steps:

1) taking the abscissa SymbolRepresents rounding down;

2) taking the corresponding ordinate as y_c=|X_l(c) I, c = k, the symbol "|" stands for X_l(c) Die length of (e), if X_l(c)=a+bj，

Then $\left|X_l\left(c\right)\right|=\sqrt{a^2+b^2};$

3)(f_c,y_c) Forming a coordinate pair;

for example: if X_l(1)=3+4j,|X_l(1) If | =5, then abscissa $f_1=\frac{f}{N_3}\&CenterDot;1,$ Ordinate of the curve $y_1=\left|X_l\left(1\right)\right|=\sqrt{3^2+4^2}=5.$

105: making a statistical histogram of the frequency spectrum distribution by using the spectrogram;

the method comprises the following steps: will [0, f/2 ] of the spectrogram]The interval is divided into M segments (the value of M can be selected according to the actual situation, and the embodiment of the invention takes 10 as an example for explanation); for an abscissa value f within each segment interval_cThe sum of their respective ordinate values, i.e. h, is calculated_m=∑y_c(ii) a Obtain a statistical histogram vector H_l=(h₁,h₂,…,h_m,...,h_M)。

106: setting a comparison interval N of packet voice data₂=βN₁β takes 2, and if it can be changed as appropriate according to the actual situation, it corresponds to the interval of the index l of the packet sound dataSubtracting the histogram vectors corresponding to the grouped sound data with the subscript difference delta l and calculating the square of the modulus to obtain a difference data sequence { G_p}；

The specific implementation process of the step is as follows: calculation of G_p=|H_p+Δl-H_p|²,p=1,2,…,N\N₃Δ l, for example: when there are 16 groups of the grouped sound data, i.e., the maximum value of l is 16, Δ l is 10,

calculation of G₁=|H₁₁-H₁|²,G₂=|H₁₂-H₂|²,…,G₆=|H₁₆-H₆|²The difference data sequence is { G }₁,G₂,…,G₆}。

107: by a sequence of difference data G_pAnd the number N of sampling points of sound data of pipeline burying transition section₁Obtaining a fitting result function and carrying out peak value detection;

the step toolThe body is as follows: retrieving a sequence of difference data G_pMaximum of G, assigning the maximum value to G_maxThe threshold coefficient (the value thereof may be selected according to actual conditions, and the embodiment of the present invention is described with 0.5 as an example). If there is continuous data G_pGreater than a threshold value G_maxThen the sequential data is organized into an array subset { G }_p}_qQ =1,2, 3.. is less than G_maxData omission is not considered. That is, the difference data series { G ] is converted by this operation_pIs divided into a plurality of sections larger than a threshold value G_maxQ is derived from a sequence of difference data { G }_pAnd a threshold value G_maxIt is determined that continuous in the embodiment of the present invention generally means that 3 or more data become continuous data.

For example: the difference data sequence is {0,0.5,2.3,5.1,7.9,6.4,3.2,1.1,0.9,1.5,5.8,7.6,9.3,10.0,6.2,1.6,0.8}, and the maximum value G is set_maxIf =10.0, the threshold value G is determined_maxAt 5.0, the difference data series is divided into 2 subsets of groups as follows:

{G_p}₁={5.1,7.9,6.4}，{G_p}₂= 5.8,7.6,9.3,10.0, 6.2. In addition, the air conditioner is provided with a fan,

for example: the difference data sequence is {0,0.5,2.3,5.1,4.9,6.4,3.2,1.1,0.9,1.5,5.8,7.6,9.3,10.0,6.2,1.6,0.8}, and the maximum value G is set_maxIf =10.0, the threshold value G is determined_maxAt 5.0, the difference data series is divided into 1 subset of the set of segments as follows:

{G_p}₁={5.8,7.6,9.3,10.0,6.2}。

for difference data sequence { G_pQ-th tuple subset of { G }_p}_qPerforming least square fitting of Gaussian function to obtain fitting result functionThe undetermined coefficients a, b and c are obtained through the following algebraic equation system:

$f\left(x\right)=\underset{p=1}{\overset{N_q}{\mathrm{\&Sigma;}}}{({\mathrm{ae}}^{-{(x_p-b)}^2/c^2}-y_p)}^2$

$\frac{\&PartialD;f}{\&PartialD;a}=2\underset{p=1}{\overset{N_q}{\mathrm{\&Sigma;}}}({\mathrm{ae}}^{-{(x_p-b)}^2/c^2}-y_p)e^{-{(x_p-b)}^2/c^2}=0$

$\frac{\&PartialD;f}{\&PartialD;b}=2a\underset{p=1}{\overset{N_q}{\mathrm{\&Sigma;}}}({\mathrm{ae}}^{-{(x_p-b)}^2/c^2}-y_p)e^{-{(x_p-b)}^2/c^2}\&times;\frac{2(x_p-b)}{c^2}=0$

$\frac{\&PartialD;f}{\&PartialD;c}=2a\underset{p=1}{\overset{N_q}{\mathrm{\&Sigma;}}}({\mathrm{ae}}^{-{(x_p-b)}^2/c^2}-y_p)e^{-{(x_p-b)}^2/c^2}\&times;\frac{2{(x_p-b)}^2}{c^3}=0$

in the formula, N_qNumber of elements, x, of array subset_p=p，y_p=G_p. As can be seen from the nature of the gaussian function, a peak is detected at t = b.

108: and judging the statistical histogram corresponding to the peak value, and further identifying the burying condition of the pipeline.

The basic principle of the judgment in the step is as follows: when the pipeline is completely buried by soil, the sound of the collision between the inner detection body and the pipeline is relatively low, namely the sound spectrum is concentrated in a relatively low-frequency area, and in this case, a peak value exists in a low-frequency area of a spectrogram; when the pipeline is exposed, the sound of the collision between the inner detection body and the pipeline is relatively crisp, namely the sound spectrum is concentrated in a relatively high-frequency area, and in this case, peaks exist in both a low-frequency area and a high-frequency area of the spectrogram. Therefore, by determining the distribution of the spectrum of the sound, the buried state of the pipe can be determined, and the spectral distribution information of the sound can be obtained from the statistical histogram. When the buried condition of the pipeline is not changed, the sound frequency spectrum signal is stable, and the difference data sequence { G_pStable and of small value; when the buried condition of the pipeline changes, the sound frequency spectrum signal changes along with the change, and the difference data sequence { G_pThe value also increases. Therefore, the peak in the above step indicates a pipe shapeReferring to fig. 1, ① shows that the frequency spectrum of the sound signal changes obviously and the buried condition of the pipeline changes, ② shows that the frequency spectrum of the sound signal is relatively stable and the buried condition of the pipeline does not change, and the steps are as follows:

1) determiningCorresponding statistical histogram vector H₁For discrete data in the vector { h }₁,h₂,…,h_MThe least squares fit of the quadratic polynomial is also performed and the peak is detected and the abscissa value/at the peak is recorded, as shown in step 107.

2) When a peak value is detected and l' < γ M, γ is a low frequency threshold coefficient (the value of γ can be flexibly selected according to the actual situation, and the embodiment of the present invention is described by taking 0.4 as an example). As shown in fig. 2, it is shown that the sound spectrum is concentrated in the low frequency region, and the pipe enters a completely buried state;

3) when two peaks are detected and l₁’<γM、l₂’>M is a high frequency threshold coefficient, and the value > γ (the value can be flexibly selected according to the actual situation, and 0.6 is taken as an example for explanation in the embodiment of the present invention). As shown in fig. 3, it is shown that the sound spectrum has a distribution in both the low frequency region and the high frequency region, and the duct enters the bare state.

Those skilled in the art will appreciate that the drawings are only schematic illustrations of preferred embodiments, and the above-described embodiments of the present invention are merely provided for description and do not represent the merits of the embodiments.

The above description is only for the purpose of illustrating the preferred embodiments of the present invention and is not to be construed as limiting the invention, and any modifications, equivalents, improvements and the like that fall within the spirit and principle of the present invention are intended to be included therein.

Claims (4)

1. A method for detecting a buried condition of a submarine pipeline, the method comprising:

using packet voice data X after fast Fourier transform_l(k) Making a spectrogram of the array subset; making a statistical histogram of the frequency spectrum distribution through a spectrogram;

setting a comparison interval of the grouped sound data, subtracting histogram vectors corresponding to the grouped sound data with subscript difference delta l and solving the square of a modulus of the histogram vectors to obtain a difference data sequence;

acquiring a fitting result function through the difference data sequence and the number of sampling points of the sound data of the pipeline burying transition section, and performing peak value detection;

judging the statistical histogram corresponding to the peak value, and further identifying the burying condition of the pipeline;

the step of obtaining the fitting result function through the difference data sequence and the number of the sampling points of the sound data of the pipeline burying transition section specifically comprises the following steps:

retrieving a maximum value in the difference data sequence, assigning the maximum value to G_maxSetting a threshold coefficient, if there is continuous data greater than a threshold G_maxThen, the continuous data is compiled into an array subset; for difference data sequence { G_pQ-th tuple subset of { G }_p}_qAnd performing least square fitting of the Gaussian function to obtain a fitting result function.

2. The method of claim 1, wherein the grouped sound data X after the fast fourier transform is used_l(k) Before the step of making the spectrogram of the array subset, the method further comprises:

fixing a sound sensor at any position in a detector in the pipeline to acquire sound data in the pipeline and transmitting the sound data to an upper computer; performing fast Fourier transform on each packet of voice data to obtain X_l(k)。

3. The method according to claim 1, wherein the step of creating the statistical histogram of the spectral distribution by using the spectrogram specifically comprises:

will [0, f/2 ] of the spectrogram]The interval is divided into M sections; for an abscissa value f within each segment interval_cCalculating the sum of the corresponding ordinate values; obtaining a statistical histogram vector; wherein f is the sampling frequency of the sound sensor.

4. The method according to claim 1, wherein the step of determining the statistical histogram corresponding to the peak value and further identifying the buried condition of the pipeline comprises:

1) determiningCorresponding statistical histogram vector H_lFor discrete data in the vector { h }₁,h₂,…,h_MPerforming least square fitting of a quadratic polynomial, detecting a peak value, and recording an abscissa value l' at the peak value; wherein,represents rounding down;

2) when a peak is detected and l' < gammam, gamma is a low frequency threshold coefficient indicating that the pipeline is entering a fully buried state;

3) when two peaks are detected and l₁’<γM、l₂’>M is a high frequency threshold coefficient, and>gamma, indicating that the pipe is exposed.