RU2626583C1 - Method for detecting and classifying changes in parameters of pipeline jacket and its environment - Google Patents

Abstract

FIELD: measuring equipment.

SUBSTANCE: in a pipe jacket, a sequence of vibroacoustic pulses is excited at intervals exceeding the correlation interval of the noises existing in it, the sequence of counts of the recorded reactions for each action at the other end of the monitored pipeline section is summed up with previously obtained analogous counts, the resulting signal module is normalized and taken as the distribution density of count time intervals from the beginning to the end of the signal generated in the summing unit, according to this distribution, its estimations of expectation, mean square deviation, asymmetry and asymmetry are calculated, by the aggregate of each of these points, the regression lines of their averages and deviations from them are determined, these lines are compared with the lines calculated in the previous step and when the comparison results reach the set values, their behavior is predicted with increased in the summary amount to ensure permissible confidence limits of the calculated points, upon reaching which presence and type of changes in the pipeline system are estimated at the present moment.

EFFECT: increased reliability of detecting changes in parameters in the pipeline system and recognizing their type.

2 cl, 8 dwg

Description

The invention relates to the safety control of operated pipelines and can be used to prevent the insertion of cuts into the pipe, ammunition for undermining it, simulators of unauthorized work in the security zone of the pipeline for disinformation of the security service, as well as detection of product leaks, freezing of the ground in the current period, subsidence, bulging, waxing pipes.

A known method of detecting changes in the state of a section of the pipeline by vibro-acoustic signals that are formed during unauthorized interaction of the subject with the shell of the pipe [Protection of pipelines from unauthorized taps. / A.A. Cossacks // Security Systems. - 2008. - No. 5. - S. 150-154]. The disadvantage of this method is the small nomenclature of the detected states (impact on the pipe), the delay in the appearance of a warning signal, eliminating the possibility of preventing a violation of the integrity of the pipe and the resulting emergency.

A known method of detecting an "emergency" section of the pipeline, based on the excitation of shock vibroacoustic pulses in the shell of the pipe using welded to it sound conductive rods, followed by determining the ratio of the resonant frequency of the diagnosed pipeline to the reference [Pat. 2350833 RF, IPC F17D 5/00. A method for monitoring and diagnosing the condition of the pipeline [Text]. / Tolstunov S.A., Moser S.P., Tolstunov A.S.]. The method is based on the well-known ratio of the dependence of the resonant frequency of the plate ƒ ₀ on its thickness h: ƒ ₀ = c / 2h, and s is the propagation velocity of longitudinal waves in the pipeline. Using this pattern to identify earthworks in a protected area is not possible.

Known application No. 2006137406/28 dated 23.10.2006 (date of publication of the application 04/27/2008) for a method and device for early detection of leaks in a pipeline. According to the application, periodic pressure waves are generated in the pumped product, which are recorded at the other end of the controlled area. The distortion of the recorded wave judges the presence of leakage in this area. The method does not allow to record changes that occur outside the shell of the pipeline.

A known method for detecting leaks in the pipeline transport of hydrocarbons, based on the registration and analysis of infrasound signals in the pumped product [Pat. US 666861982. Pattern matching for real time leak detection and location in pipelines] and in various versions described in http://acoustic-solution-intl.com/faq_index .htm; www.grouplb.com; http://torinsk.ru/publication/25-mpp2007.html and others. The disadvantage of this method is the fact of violation of the integrity of the pipeline, and not preparatory work on the installation of a tie-in or ammunition.

A known method of changing environmental parameters in the environment of a buried trunk pipeline [US Pat. 2463590 RF, IPC G01N 29/04. A method for detecting changes in environmental parameters in the environment of a buried trunk pipeline. / Epifantsev B.N., Fedotov A.A.]. According to the method, vibroacoustic signals are excited in the pipe shell, recorded at a distance from the place of excitation, registered signals corresponding to the particular state of the test object are accumulated, and the accumulation result is taken as the corresponding standard with which the accumulated signals are compared during the monitoring periods. The disadvantage of this method is the unknown number of accumulations during training and testing operations, it depends on changes in weather conditions, product pumping mode, etc.

Of the known technical solutions, the closest in combination of essential features to the claimed is a method for detecting defects in pipelines [Pat. 2439551 RF, IPC G01N 29/04. A method for detecting defects in pipelines. / Alekseev S.P. and etc.].

According to the method, acoustic signals are generated and recorded after passing through the controlled medium, the distribution density of the recorded signals is determined, the moments of the obtained distributions are calculated, the magnitude of which is used to judge the presence of changes in the controlled medium. The disadvantage of this method is the identification of only one change from the normatively established. In addition, the method is operable with a high signal to noise ratio. In reality, this condition is rarely fulfilled, and the reliability of detecting changes in the environment (the presence of an inclusion) is insufficient.

The aim of the invention is to increase the reliability of detection of significant changes in the parameters of the shell of the pipeline and its environment with an increase in the range of classified changes.

This goal is achieved by the fact that vibroacoustic signals are excited in the pipe shell by successive influences on its surface at intervals exceeding the correlation interval of vibroacoustic noises existing in it, the sequence of counts of recorded reactions to each impact on the other end of the controlled section of the pipeline is summed with previously obtained similar samples, the module of the resulting signal is normalized and taken as the distribution density of the time intervals of samples about t of the beginning to the end of the signal generated in the adder, this distribution calculates its estimates of the mathematical expectation, standard deviation, asymmetry and excess, and the totality of each of these moments obtained after the next exposure to the pipeline determines the regression lines of their averages and deviations from them, compare these lines with those calculated in the previous step, and when the comparison results are below the set values, their behavior is predicted with an increase in the number of summations to ensure Permissible confidence limits calculated moments at which decide how the presence and form of changes in the pipeline system at the current time.

In addition, in the intervals between the effects on the pipe shell, the vibro-acoustic noise existing in it is recorded, its correlation interval is determined by which the intervals between the vibrations of the vibro-acoustic signals in the pipe shell are set.

The invention is illustrated by the following description and the accompanying drawings.

In FIG. 1 shows a structural diagram of a device that implements the proposed method.

FIG. 2 and 3 explain the signal generation algorithm in the adder.

In FIG. Figures 2a and 3a show noise realizations obtained by modeling a random process with an exponential correlation function and a correlation interval of 100 ms. On the first of them, a signal in the form of a cosine pulse is additively put, on the second - in the form of a log-normal function. The signal-to-noise ratio in both cases is 0.5. FIG. 2b and 3b reflect the results of a 50-fold summation of different implementations of the form shown in FIG. 2a and 3a. Similar functions are shown in FIG. 2c and 3c, obtained by 100 summations of such implementations.

In FIG. 4 illustrates changes in the values of the moment of distribution density of the kind shown in FIG. 3, and its confidence limits on the number of summations (effects on the object of control).

- средняя длительность формирующихся на выходе приемника сигналов, Л₂ - линия регрессии момента распределения, Л₁ и Л₃ - изменения доверительных границ вычисляемого момента в зависимости от количества суммирований, N_y - количество суммирований, при котором вычисляемые линии регрессии Л₁-Л₃ на предыдущем шаге N=N_y-1 отличались от полученных на заранее установленную величину, Д₁-Д₂ - доверительные интервалы, допускаемые при оценке соответствующего момента распределения, штриховые кривые - прогноз доверительных границ с увеличением N.Designations in the drawings: 1 - image of the pipeline; B is an element for exciting vibroacoustic signals in a pipe shell; G - signal generator, P - receiver of vibro-acoustic signals, U - amplifier, LZ - delay line, KL ₁ and KL ₂ - electronic keys, B ₁ and B ₂ - calculators, KOM - switch, N - number of summations, SNR - signal ratio / noise, u (t) is the normalized signal amplitude in the interval tt _k + τ _s , τ _s is the duration of the excitation of the pipe shell surface,

is the average duration of the signals formed at the output of the receiver, L ₂ is the regression line of the distribution moment, L ₁ and L ₃ are changes in the confidence limits of the calculated moment depending on the number of summations, N _y is the number of summations at which the calculated regression lines L ₁ -L ₃ at the previous step, N = N _y -1 differed from those obtained by a predetermined value, D ₁ -D ₂ - confidence intervals allowed when assessing the corresponding moment of distribution, dashed curves - forecast of confidence boundaries with increasing N.

When element B acts on the pipe shell, a vibroacoustic pulse is formed in it, propagating in the direction of receiver P. Depending on the ratio of the wave impedances of the contacting systems “shell - pumped product”, “shell - external medium”, part of the pulse energy goes into these media, and vibroacoustic vibrations that have left the envelope depend on the frequency of their vibrations [S. Budenkov et al. Evaluation of the capabilities of the acoustic emission method for monitoring pipelines // Defectoscopy, 2000, No. 2, P. 29-36]. For these reasons, the shape of the pulse formed at the site of interaction of element B with the surface of the pipe, as it propagates, varies depending on the amplitude-frequency characteristics of the vibration propagation channel. Waxing of the inner wall of the pipe changes the wave impedance ratio “pipe sheath - pumped product”, not only the pulse energy changes, but also its shape due to the dependence of the wave resistances on the oscillation frequency [Non-destructive testing: Reference: V 8 t. / Under total. ed. V.V. Klyueva. T. 3. I.N. Ermolov, Yu.V. Lange. Ultrasonic inspection - M.: Mechanical Engineering, 2006. - 864 p.].

Random fluctuations of mechanical (vibroacoustic) vibrations are generated in the pipeline shell caused by the following reasons [B. Epifantsev et al. Pipeline transport: neutralization of new security threats. - Omsk: SibADI, 2006. - 295 p.]:

- the turbulent nature of the movement of the pumped product;

- a wind factor forming seismic vibrations through the vegetation in the transmission channel of the “ringing” vibroacoustic pulses;

- transport passing near the “security zone”, etc.

The signal registered by the receiver P turns out to be comparable or less than the squared deviation of the accompanying noise. In this regard, to detect incoming signals, it is necessary to introduce an operation to increase the signal-to-noise ratio. It is proposed to use signal accumulation as such an operation. The problem under study admits this possibility - the decision is made by the totality of responses to N “ringing” vibroacoustic pulses.

The electric impulse formed by the generator G is converted by the excitation element B into a vibroacoustic, which “rings” the channel of its propagation. Having reached the receiver P, it is again converted into an electric signal coming through the amplifier to the delay line LZ. At the end of the signal generator G switch KOM opens the keys KL ₂ , formed during the operation of the generator, the response at the output of the receiver P is transferred to the computer В ₂ . After the end of the pulse of generator G, a noise process enters the delay line LZ, the intensity distribution of which in the interval between “ringing” is transferred by the KOM switch to the computer B ₁ before the appearance of a new pulse of generator G.

. Этот эффект справедлив, когда реализации шумов не коррелированны. Для обеспечения этого условия в вычислителе B₁ определяется интервал корреляции текущих шумов τ_k и по нему устанавливается требуемая длительность импульса τ_с. Наилучший случай - равенство τ_с=τ_k. Система обнаружения оказывается адаптированной к изменениям шумовой обстановки в зоне контролируемого объекта.In the calculator B _2, each incoming implementation of the response from the delay line LZ is summed up with the previously recorded ones. In the first of them N = 1, noise is mainly visible, a useful signal is not allocated (Figs. 2a, 3a). After many summations (accumulation process), the signal appears more clearly. According to the principle of accumulation, the increase in signal intensity is proportional to the number of accumulations N, and for noise,

. This effect is true when noise implementations are not correlated. To ensure this condition, the correlation interval of the current noise τ _k is determined in the calculator B ₁ and the required pulse duration τ _{s is established from it} . The best case is the equality τ _c = τ _k . The detection system is adapted to changes in the noise environment in the area of the controlled object.

As the number of summations increases, the form of the received signal begins to be detected (Fig. 2, 3), by which one can judge the nature of the change in the propagation channel of the “ringing” signal [Epifantsev B.N. et al. On the sensitivity assessment of a vibroacoustic system for detecting local disturbances in environmental parameters in the environment of a main pipeline // Defectoscopy. - 2015. - No. 2. - S. 17-26].

The description of this form is possible by the moments of the forming curve. Such moments can be mean, standard deviation, asymmetry, excess. Performing this operation requires knowledge of the probability density function. If we take the duration from the start of the generator operation at the moment t-τ _k to the next signal sample (the number of samples is determined by the selected number of KL ₂ keys) as the distribution parameter, then the amplitudes of the samples of the next sum can be taken as probability values. For this, it is necessary to use the signal modulus u (t) (the probabilities cannot be negative, this effect is observed at the beginning of the operation of the calculator (for N = 1)). The second condition is reduced to the normalization of the values of the samples (their sum should be equal to one).

Finally, the question should be solved when the process of summing up the responses of the pipeline laying channel to “ringing” signals is completed. FIG. 4 explains the algorithm for finding the answer to this question.

If N is small, the scatter of the calculated moments according to the above scheme is significant. With increasing N, the moment estimates are grouped near the average (the variance of the estimates is proportional to the root of 1 / N). After each summation, the regression lines of the estimated parameters and deviations from them (L ₁ -L ₃ ) are determined. If during the next calculation of these lines, their difference from the previous ones does not exceed the specified value, their behavior is predicted with increasing N (dashed lines in Fig. 4). It is further determined at which value of N the found estimate falls within a given confidence interval (D ₁ -D ₂ in Fig. 4). The implementation of this number of “ringing pulses” completes the cycle of the next test of the pipeline system for any changes that have occurred.

The list of changes of interest is being implemented at current landfills. For each simulated change, the response form is evaluated for different N and the position of the shape moments with confidence intervals.

Claims (2)

1. A method for detecting and classifying changes in the pipeline shell and its environment, based on the excitation and registration of vibroacoustic signals, forming the probability density of the recorded signals and calculating its moments, deciding on their values about the appearance of changes in the control object, characterized in that vibroacoustic signals excite in the pipe shell by successive influences on its surface at intervals exceeding the existing correlation interval it contains vibro-acoustic noise, the sequence of samples of the recorded reactions to each impact at the other end of the pipeline section being monitored is summed with the previously obtained similar samples, the resulting signal module is normalized and taken as the density of the distribution of the time intervals of the samples from the beginning to the end of the signal generated in the adder, this distribution is calculated his estimates of mathematical expectation, standard deviation, asymmetry and kurtosis, in the aggregate of each of these x moments determine the regression lines of their means and deviations from them, compare these lines with those calculated in the previous step, and when the results of comparing the set values are achieved, they predict their behavior with an increase in the number of summations to ensure acceptable confidence limits for the calculated moments, which are judged as being available, as well as the form of changes in the pipeline system at the current time. 2. The method according to p. 1, characterized in that in the intervals between the effects on the shell of the pipeline register the vibro-acoustic noise existing in it, determine its correlation interval, which sets the intervals between the excitations of vibro-acoustic signals in the pipe shell.

Images