KR102568086B1 - Apparatus and Method for Detecting Leak of Pipe - Google Patents

Abstract

The present invention uses an Acoustic Emission sensor ("AE sensor") to detect an acoustic emission signal generated in a fluid transport pipe and measures its change, and at the same time uses a "vibration sensor" to detect the acoustic emission signal generated in a fluid transport pipe. "A device for detecting leaks in a fluid transport pipe" and "a device for detecting leaks in a fluid transport pipe" that detects and diagnoses whether or not there is a leak in a fluid transport pipe (fluid leak) by detecting and measuring the degree of vibration and measuring the degree of change of such vibration. detection method".

Description

Apparatus and Method for Detecting Leak of Pipe by Measuring Acoustic Emission Signal and Vibration Acceleration {Apparatus and Method for Detecting Leak of Pipe}

The present invention relates to a technology capable of quickly detecting or diagnosing leaks in various types of fluid transport pipes, including double thermal insulation pipes for transporting a fluid or gas heat medium such as high-temperature hot water. Specifically, a sensor ("AE sensor") for detecting an acoustic emission signal is used to detect an acoustic emission signal generated in a fluid transport pipe and to measure its change, and at the same time, a vibration sensor is used to "Detecting and diagnosing whether there is a leak in the fluid transport pipe (fluid leak) by detecting the vibration of the fluid transport pipe, measuring the degree (vibration acceleration or equivalent physical quantity), and further measuring the degree of change of this vibration" It relates to a leak detection device of a fluid transport pipe" and a "leak detection method of a fluid transport pipe".

A heat transport pipe that transports a fluid or gaseous heat medium at a predetermined pressure is generally a double insulation pipe (pre-insulated pipe) composed of an inner tube, an insulation material (intermediate layer), and an outer tube. this is used Because these heat transport pipes are buried underground, it is difficult to maintain with the naked eye. In particular, even if a leak occurs in a heat transport pipe due to breakage, crack, etc., it is very difficult to accurately determine the fact and location. Accordingly, various types of methods have been proposed to detect or diagnose leaks in heat transport pipes. Korean Patent Registration No. 10-1447920 and the like suggest a technique for determining whether an underground pipe is leaking by using a leak sensor. Japanese Patent No. 6331164 discloses a technique of detecting a leak location by detecting the leaked hydrogen gas by exploring the ground with a hydrogen gas explorer in a state in which hydrogen is filled in the pipe of a water supply pipe and pumped.

However, in the case of the conventional technology for detecting or diagnosing a leak in a fluid transport pipe including a heat transport pipe, the reliability of the detection is low, and it takes considerable time to calculate and analyze the signal from the sensor. There is a limit to detecting or diagnosing pipe leaks in real time.

Republic of Korea Patent Registration No. 10-1447920 (2014. 10. 08. notice). Japanese Patent No. 6331164 (published on May 11, 2018).

The present invention was developed to overcome the limitations of the prior art as described above, and in detecting or diagnosing a leak in a fluid transport pipe including a heat transport pipe, it is possible to increase the reliability of detection and diagnosis, and for such detection and diagnosis An object of the present invention is to provide a technology capable of detecting and diagnosing a leak in a fluid transport pipe in real time and with high accuracy and reliability by reducing the time required.

In order to achieve the above object, in the present invention, an inlet AE sensor and an outlet AE sensor installed at each of the inlet and outlet of the fluid transport pipe to detect and measure the acoustic emission signal generated in the fluid transport pipe; an inlet vibration sensor and an outlet vibration sensor installed at each of the inlet and outlet parts of the fluid transport pipe and detecting and measuring a vibration signal of the fluid transport pipe generated by the elastic motion of the fluid transport pipe; And a signal analysis control device for receiving signals from the AE sensor and the vibration sensor to determine whether or not the fluid transport pipe is leaking; The signal analysis control device performs filtering on the original signal received from the AE sensor and the vibration sensor, calculates the spectrum of the original signal and the filtered signal, and calculates the statistical feature value from the spectrum of the original signal and the filtered signal. is calculated, and the calculated statistical feature value is compared with each preset reference value, and when the calculated statistical feature value is within the allowable range of the preset reference value, it is determined as a "normal state", and when it is out of this range, It is judged as "abnormal state", and if all calculated statistical feature values are judged to be "abnormal state", it is determined as <leakage state>, and if any one is judged to be "normal state" in the comparison result, it is judged as <leakage judgment pending> An apparatus for detecting and diagnosing leakage of a fluid transport pipe, characterized in that to be provided.

In addition, in the present invention, in order to achieve the above object, a method of detecting and diagnosing a leak in a fluid transport pipe using the leak detection device of the present invention is provided.

In the present invention, the AE sensor for detecting the acoustic emission signal is used to detect the acoustic emission signal generated in the fluid transport pipe and the change is measured, and at the same time, the vibration sensor is used to detect the vibration of the fluid transport pipe and measure the degree It measures the degree of change of these vibrations, and simultaneously uses these two measurement signals to detect and diagnose whether or not there is a leak in the fluid transport pipe (fluid leak), so the reliability and accuracy of the detection and diagnosis results are very high. It has the advantage of being high.

In particular, according to the present invention, since the detection and diagnosis of whether or not there is a leak in the fluid transport pipe is performed very quickly, it is possible to grasp the state of the fluid transport pipe in real time in the field, and accordingly, an immediate response is possible. The effect of being able to prevent various problems and difficult situations that may occur due to water leakage is exhibited.

1 is a schematic view conceptually showing that the leak detection device of the present invention is installed in a fluid transport pipe.
Figure 2 is a schematic block diagram showing the configuration of the signal analysis control device provided in the leak detection device of the present invention.
Figure 3 is a schematic flow chart showing a series of processes in the leak detection method of the present invention.

Hereinafter, preferred embodiments of the present invention will be described with reference to the accompanying drawings. The present invention has been described with reference to the embodiments shown in the drawings, but this is described as one embodiment, and thereby the technical idea of the present invention and its core configuration and operation are not limited.

1 is a schematic diagram conceptually showing that the leak detection device 100 of the present invention is installed in the fluid transport pipe 200. As illustrated in the drawing, the leak detection device 100 according to the present invention is installed at two spaced apart points of the fluid transport pipe 200, that is, the inlet and the outlet, respectively, and emits acoustic waves generated from the fluid transport pipe. Emission) generated by the AE sensors (1a, 1b) that detect and measure the signal and the elastic movement of the fluid transport pipe 200 according to the flow of the fluid installed at the inlet and outlet of the fluid transport pipe 200 It is configured to include vibration sensors (2a, 2b) for detecting and measuring the vibration signal of the fluid transport pipe, and a signal analysis control device (3) for receiving a signal from the sensor and determining whether or not there is a leak in the fluid transport pipe. The "inlet" becomes the starting point of a section (leakage examination section) for determining whether or not there is a leak in the fluid transport pipe 200, and the "outflow section" is the end point of the leak checking section corresponds to

When the fluid flows in a state where pressure is applied to the inside of the fluid transport pipe, an acoustic signal is generated due to the elasticity of the fluid transport pipe. Such an acoustic signal generated in the fluid transport pipe (abbreviated as "acoustic emission signal") varies depending on whether there is a leak in the fluid transport pipe. That is, there is a difference between an acoustic emission signal in a “normal state” in which there is no leakage in the fluid transport pipe and an acoustic emission signal in a “leakage state” in which leakage occurs.

In the present invention, an AE sensor capable of detecting and measuring an acoustic emission signal is installed at two points, that is, an inlet and an outlet, respectively, separated by a predetermined distance from the fluid transport pipe 200, so that the sound measured by the AE sensor By observing the change of the emission signal, it is possible to determine whether the fluid transport pipe is leaking. To this end, the leak detection device 100 of the present invention has an "inlet AE sensor 1a" installed at the inlet of the fluid transport pipe 200 and an "outlet AE sensor 1b" installed at the outlet, there is. Therefore, at the inlet corresponding to the starting point of the leak check section, the acoustic emission signal generated at the inlet is measured by the inlet AE sensor 1a, and the end point corresponding to the end point of the leak check section In the outlet, the acoustic emission signal generated in the outlet is measured by the outlet AE sensor 1b.

On the other hand, the turbulent flow of the fluid flowing in a state in which pressure is applied inside the fluid transport pipe causes energy waves, and as a result, soft vibrations occur between the fluid and the fluid transport pipe. However, when a leak occurs in the fluid transport pipe, the fluid passes through the defective portion (mainly a crack/crack) of the fluid transport pipe and causes friction with the defective portion, thereby vibrating the fluid transport pipe. The vibration of the fluid transport pipe due to such a water leak shows a different aspect from the vibration that occurs in the fluid transport pipe when there is no leak. The signal related to such vibration may be vibration acceleration or a signal expressed by another physical quantity equivalent to other vibration acceleration. Therefore, in the present invention, a signal indicating such vibration acceleration and a physical quantity equivalent thereto is collectively referred to as a "vibration signal".

In the present invention, vibration sensors 2a and 2b are additionally installed at the inlet and outlet of the fluid transport pipe 200 to detect and measure vibration signals generated in the fluid transport pipe 200, and the vibration sensors 2a , By observing the change in the vibration signal measured in 2b), it is used to determine whether there is leakage in the fluid transport pipe. Specifically, the leak detection device 100 of the present invention is provided with an "inlet vibration sensor 2a" installed at the inlet of the fluid transport pipe 200 and an "outlet vibration sensor 2b" installed at the outlet, there is. Therefore, at the inlet corresponding to the starting point of the leak checking section, the vibration signal generated at the inlet is measured by the inlet vibration sensor 2a, and the outflow corresponding to the end point of the leak checking section In the outlet, a vibration signal generated in the outlet is measured by the outlet vibration sensor 2b.

If the fluid transport pipe is a double insulation pipe having an inner tube and an outer tube, and it is necessary to determine whether there is a leak in the inner tube, the inner tube is artificially exposed when the AE sensors 1a and 1b and the vibration sensors 2a and 2b are installed. Alternatively, the part where the inner pipe is naturally exposed or the valve at the pipe inspection port can be used as an installation point.

2 is a schematic block diagram showing the configuration of the signal analysis control device 3 provided in the water leak detection device 100 of the present invention. As illustrated in the drawing, the signal analysis control device 3 of the present invention receives signals from the AE sensor and the vibration sensor to determine whether the fluid transport pipe is leaking. Specifically, the signal analysis control device 3 is composed of a signal collection/storage module 31, a filter module 32, a spectrum conversion module 33, a statistical characteristic value calculation module 34, a water leakage determination module 35, and an output module 36. The signal analysis control device 3 may be composed of a processor of a general-purpose computer, a special-purpose computer, or other programmable data processing equipment equipped with computer program instructions (execution engine). In particular, each module constituting the signal analysis control device 3 means a unit that processes each unique function or operation, which may be implemented as hardware or software, or a combination of hardware and software.

3 is a schematic flow chart of a leak detection method of a fluid transport pipe according to the present invention. Referring to FIGS. 2 and 3, looking at the operation performed by each module constituting the signal analysis control device 3 in detail, in the signal collection / storage module 31, the inlet AE sensor 1a, the outlet Collects and stores the original signals (raw data / "raw signal") measured by each of the AE sensor 1b, the inlet vibration sensor 2a, and the outlet vibration sensor 2b in real time in a wired/wireless manner (Step S1).

The filter module 32 generates a filtered signal by applying a filter to the collected original signal (step S2). If the above filtering operation is performed on the original signals measured by the AE sensor and the vibration sensor, only frequencies showing characteristic changes that can identify the normal state and leak state of the fluid transport pipe can be emphasized. At this time, the filter module 32 can generate a filtered signal by selecting or applying all of a pass filter, a high pass filter, and a low pass filter as a filter, and using all of the above three filters provides better classification performance. Therefore, it is preferable to filter the original signal using all of a bandpass filter, a highpass filter, and a lowpass filter.

The spectrum conversion module 33 calculates a spectrum by performing spectrum conversion on each of the original signal and the filtered signal (step S3). That is, the "spectrum of the original signal" and the "spectrum of the filtered signal" are respectively calculated. When calculating the spectrum, a known calculation method such as a method using a Fourier transform or the like can be used.

In the statistical feature value calculation module 34, the following "statistical feature values" are calculated from the calculated "spectrum of the original signal" and "spectrum of the filtered signal" (step S4).

Specifically, in the statistical feature value calculation module 34, <frequency skewness>, <frequency skewness>, <frequency skewness>, < Calculate three statistical feature values: frequency center> and <normal negative log-likelihood>.

일 수 있다. Here, <Frequency skewness> is a statistical characteristic value representing the degree of asymmetry of the spectrum, and according to Equation 1 below, the spectrum of a signal (high-pass filtered signal) created by applying a high-pass filter to the original signal Statistical feature values calculated

can be

일 수 있다. <Frequency center> means the frequency at the center of the spectrum, and is a statistic calculated according to Equation 2 below on the spectrum of a signal (low-pass filtered signal) created by applying a low-pass filter to the original signal. feature value

can be

이다. <Normal negative log-likelihood> is a value obtained by taking a negative logarithm of the probability of a normalized signal and summing the frequencies in all ranges. Statistical feature values created by applying Equation 3 below to the signal)

am.

는 주파수 왜도를 의미하며,

는 주파수 중심을 의미하고,

은 정규 음성 로그우드를 의미한다. 그리고

은 샘플링된 신호의 수를 의미하며,

는 주파수를 의미하고,

는 스펙트럼 산술평균을 의미한다. 그리고

는 해당 주파수의 스펙트럼을 의미하고,

은 샘플링 레이트를 의미한다. 그리고

는 음향 방출 신호 또는 가속도 신호를 의미하며,

는 신호의 산술평균,

는 신호의 표준편차, 그리고

는 스펙트럼의 표준편차를 의미한다. In Equation 1, Equation 2 and Equation 3 above,

is the frequency skewness,

is the frequency center,

means regular voice logwood. and

is the number of sampled signals,

means frequency,

means the spectral arithmetic average. and

denotes the spectrum of the corresponding frequency,

is the sampling rate. and

Means an acoustic emission signal or an acceleration signal,

is the arithmetic mean of the signal,

is the standard deviation of the signal, and

is the standard deviation of the spectrum.

On the other hand, in the statistical feature value calculation module 34, <regular voice logwood (regular voice logwood ( Calculate three statistical feature values: Normal negative log-likelihood>, <Variance frequency>, and <Standard deviation>.

이다. Here, <Normal negative log-likelihood> is a statistical feature value created by applying Equation 3 to the spectrum of a signal (high-pass filtered signal) created by applying a high-pass filter to the original signal as described above.

am.

일 수 있다. <Variance frequency> represents the variance of the spectrum, and is a statistical feature value created by applying Equation 4 below to the spectrum of a signal (low-pass filtered signal) created by applying a low-pass filter to the original signal.

can be

일 수 있다. <Standard deviation> represents the standard deviation of the signal, and is a statistical feature value created by applying Equation 5 below to a signal (high-pass filtered signal) created by applying a low-pass filter to the original signal.

can be

는 분산 주파수를 의미하며,

는 표준편차를 의미한다. 그리고 앞서 설명한 것처럼

는 주파수 중심을 의미하고,

은 샘플링된 신호의 수를 의미하며,

는 주파수를 의미하고,

는 해당 주파수의 스펙트럼을 의미하며,

는 음향 방출 신호 또는 가속도 신호를 의미하고,

는 신호의 산술평균을 의미한다. In Equation 4 and Equation 5 above,

is the dispersion frequency,

means standard deviation. and as explained earlier

is the frequency center,

is the number of sampled signals,

means frequency,

denotes the spectrum of the corresponding frequency,

Means an acoustic emission signal or an acceleration signal,

means the arithmetic mean of the signal.

As described above, in the present invention, the vibration signal generated in the fluid transport pipe at the inlet is measured by the inlet vibration sensor 2a, and the vibration signal generated in the fluid transport pipe at the outlet is measured by the outlet vibration sensor 2b. to measure Therefore, the statistical feature value calculation module 34 calculates the following statistical feature values using the spectrum of the measurement signal of the vibration signal as well as the sound emission signal.

First, in the statistical feature value calculation module 34, <standard deviation>, Three statistical feature values of <Root variation frequency> and <Frequency RS overall> are calculated.

으로서, 통계 특징값 연산 모듈(34)에서는 유입부 진동 센서(2a)에 의해 측정된 진동 신호의 스펙트럼을 이용하여<표준 편차(Standard deviation)>를 산출하게 된다. <Standard deviation> related to the vibration signal is a statistical feature value created by applying Equation 5 above to the unfiltered original signal

As , the statistical characteristic value calculation module 34 calculates <Standard deviation> using the spectrum of the vibration signal measured by the inlet vibration sensor 2a.

일 수 있다. <Root variation frequency> related to the vibration signal is a value corresponding to the 1/2 power of the dispersion frequency, and is a statistical feature value created by applying Equation 6 below to the spectrum of the original unfiltered signal

can be

일 수 있다. <Frequency RS overall> related to the vibration signal is a value corresponding to the square root of the sum of the squares of the spectrum, and is a statistical feature value created by applying Equation 7 below to the spectrum of the original unfiltered signal

can be

In addition, the statistical feature value calculation module 34 uses the spectrum of the vibration signal measured at the outlet, that is, the spectrum of the vibration signal at the outlet measured by the outlet vibration sensor 2b to obtain <Frequency skewness> , <Frequency RMS> and <Normal negative log-likelihood> are calculated.

<Frequency skewness> has already been described above, so repeated description thereof will be omitted.

In the case of <Normal negative log-likelihood>, the sound emission signal was calculated by applying a high-pass filter, but the vibration acceleration signal measured at the outlet was applied to a signal that passed a low-pass filter. There is a difference in application.

일 수 있다. <Frequency RMS> is the square root of the average of the squares of the spectrum, and is a statistical feature value created by applying Equation 8 below to the spectrum of the unfiltered original signal

can be

는 주파수의 제곱평균 제곱근(Frequency RMS)이고,

는 주파수 제곱 합의 제곱근(Frequency RS overall)이며,

는 분산 주파수 제곱근(Root variation frequency)이고,

는 주파수의 분산이며,

는 주파수를 의미하고,

는 해당 주파수의 스펙트럼을 의미하며,In Equation 6, Equation 7 and Equation 8 above,

is the root mean square root of the frequency (Frequency RMS),

is the square root of the sum of squared frequencies (Frequency RS overall),

is the root variation frequency,

is the variance of the frequency,

means frequency,

denotes the spectrum of the corresponding frequency,

As described above, in the statistical feature value calculation module 34, “statistical feature values using the spectrum of the acoustic emission signal at the inlet”, “statistical feature values using the spectrum of the acoustic emission signal at the outlet”, and “statistical feature values using the spectrum of the acoustic emission signal at the inlet” When the "statistical feature value using the spectrum of the vibration signal" and the "statistical feature value using the spectrum of the vibration signal at the outlet" are calculated, respectively, the water leakage determination module 35 subsequently converts the calculated statistical feature value to each It is compared with a preset reference value to determine whether the fluid transport pipe 200 leaks (step S5 in the method of the present invention).

In a "normal state" where no leakage occurs in the fluid transport pipe, reference values for each of the statistical characteristic values described above at the inlet and outlet are preset by preliminary investigation.

In the present invention, it is determined whether leakage occurs in the fluid transport pipe 200 by comparing the statistical feature values calculated for each of the inlet and outlet by the method described above with each preset reference value.

Specifically, <Frequency skewness>, <Frequency center> and <Normal negative log calculated using the spectrum of the acoustic emission signal measured by the inlet AE sensor 1a -likelihood)> are compared with the preset reference values.

And <Normal negative log-likelihood>, <Variance frequency> and <Standard deviation calculated using the spectrum of the acoustic emission signal from the outlet AE sensor 1b The three statistical feature values of > are compared with preset reference values.

In addition, <Standard deviation> calculated using the spectrum of the vibration signal measured by the inlet vibration sensor 2a, <Root variation frequency>, and <Frequency RS <Frequency skewness> calculated using the spectrum of the vibration signal at the outlet measured by the outlet vibration sensor 2b by comparing three statistical feature values of < overall)> with a preset reference value , <Frequency RMS>, and <Normal negative log-likelihood> are compared with preset reference values.

라고 하고 이에 대한 표준편차를

라고 할 때, <

>가 된다. 즉, 사전 설정 기준값의 허용범위는 <

>가 되는 것이고, 통계 특징값이 <

> 내에 있게 될 경우에는 "정상 상태"로 판정하게 되고, 이 범위를 벗어날 경우에는 "비정상 상태"로 판단하는 것이다. 이러한 사전 설정 기준값의 허용범위는, 통계 특징값이 정규분포의 형태로 분포할 경우 통계적으로 정확도 99.7%를 보이는 것이다. In comparing each statistical feature value with a preset reference value, if the statistical feature value is within the allowable range of the preset reference value, it is determined as "normal state", and if it is out of this range, it is determined as "abnormal state". judged by At this time, the allowable range of the preset reference value is the average of the preset reference values.

and the standard deviation for this

When saying, <

> becomes That is, the allowable range of the preset reference value is <

>, and the statistical feature value is <

If it is within >, it is judged to be a "normal state", and if it is out of this range, it is judged to be an "abnormal state". The allowable range of these preset reference values is to show a statistical accuracy of 99.7% when statistical feature values are distributed in the form of a normal distribution.

As a result of judging according to the above criteria, three statistical characteristic values based on the spectrum of "acoustic emission signal" in <inlet> and three statistics based on the spectrum of "acoustic emission signal" in <outlet> Characteristic values, and three statistical feature values based on the spectrum of "vibration signal" in <inlet> and three statistical feature values based on the spectrum of "vibration signal" in <outlet>, respectively. If all of the comparison results of the preset reference values are judged to be "abnormal state", it is determined to be <leakage state>, and if any one of the comparison results is determined to be "normal state", it is determined to be <leakage judgment suspended>.

Any of the judgment results made by the water leakage determination module 35, that is, the <water leakage state> determination and the <leakage determination suspended> determination, are visualized through the output module 36 (for example, through a computer monitor or mobile terminal). It is output in various ways such as displaying the result) or auditory method (eg, displaying a warning sound through an alarm, etc.) and delivered to the manager. If the manager is notified of the <leakage condition>, he/she will take immediate countermeasures against the leak. By treating it as a necessary object, it is possible to efficiently manage the fluid transport pipe.

As described above, in the present invention, the acoustic emission signal generated from the fluid transport pipe 200 and the vibration signal generated by the elastic motion of the fluid transport pipe 200 are simultaneously used to determine whether the fluid transport pipe 200 leaks. Since it detects and diagnoses, it has the advantage that the reliability of detection and diagnosis results is very high.

3: signal analysis control device
31: signal acquisition/storage module
32: filter module
33: spectrum conversion module
34: statistical feature value calculation module
35: water leakage determination module
36: output module
200: fluid transport pipe

Claims (4)

An inlet AE sensor 1a and an outlet AE sensor 1b installed at the inlet and outlet of the fluid transport pipe 200 to detect and measure acoustic emission signals generated in the fluid transport pipe;
An inlet vibration sensor 2a installed at each of the inlet and outlet of the fluid transport pipe 200 and detecting and measuring the vibration signal of the fluid transport pipe generated by the elastic motion of the fluid transport pipe 200 and outlet vibration sensor 2b; and
Includes a signal analysis control device (3) for receiving signals from the AE sensors (1a, 1b) and the vibration sensors (2a, 2b) to determine whether the fluid transport pipe is leaking;
The signal analysis control device 3 converts the original signal measured from each of the inlet AE sensor 1a, the outlet AE sensor 1b, the inlet vibration sensor 2a, and the outlet vibration sensor 2b to A signal collection/storage module 31 that collects and stores signals in real time in a wired/wireless manner; a spectrum conversion module 33 that calculates a spectrum by performing a spectrum conversion operation on each of the original signal and the filtered signal; and three statistical characteristics of frequency skewness, frequency center, and normal negative log-likelihood using the spectrum of the acoustic emission signal measured by the inlet AE sensor 1a. Calculate the value, and use the spectrum of the acoustic emission signal at the outlet measured by the outlet AE sensor 1b to calculate the normal negative log-likelihood, variance frequency, and standard deviation ( standard deviation), and using the vibration signal measured by the inlet vibration sensor 2a and its spectrum, standard deviation, root variation frequency and frequency Three statistical feature values of the square root of the sum of squares (Frequency RS overall) are calculated, and frequency skewness (frequency skewness) and frequency It is configured to include a statistical feature value calculation module 34 that calculates three statistical feature values of frequency RMS and normal negative log-likelihood;
The signal analysis control device 3 performs filtering on the original signals received from the AE sensors 1a and 1b and the vibration sensors 2a and 2b, calculates the spectrum of the original signal and the filtered signal, and then Statistical feature values are calculated from the spectrum of the signal and the spectrum of the filtered signal, and the calculated statistical feature values are compared with respective preset reference values. If it is out of this range, it is judged to be "abnormal state", and if all calculated statistical feature values are judged to be "abnormal state", it is determined to be <leak state>, and any one of the comparison results is judged to be "abnormal state". If it is judged as "normal state", the leak detection device of the fluid transport pipe, characterized in that it is determined as <leakage judgment suspended>.
delete An inlet AE sensor (1a) and an outlet AE sensor (1b) that detect and measure the acoustic emission signal generated in the fluid transport pipe 200 are installed at the inlet and outlet of the fluid transport pipe 200, respectively. and an inlet vibration sensor 2a and an outlet vibration sensor 2b, which detect and measure the vibration signal of the fluid transport pipe generated by the elastic motion of the transport pipe 200, are installed at the inlet of the fluid transport pipe 200. And installed in each outlet, receiving and storing the signals from the AE sensors (1a, 1b) and the vibration sensors (2a, 2b) in the signal analysis control device (3);
Performing filtering on the original signals received from the AE sensors (1a, 1b) and the vibration sensors (2a, 2b) in the signal analysis control device (3);
Calculating the spectrum of the original signal and the filtered signal in the signal analysis control device (3);
calculating statistical feature values from the spectrum of the original signal and the spectrum of the filtered signal in the signal analysis control device 3; and
The calculated statistical feature value is compared with each preset reference value, and if the calculated statistical feature value is within the allowable range of the preset reference value, it is judged as "normal state", and if it is out of this range, it is judged to be "abnormal state". , and if all calculated statistical feature values are judged to be "abnormal state", it is determined as <leakage state>, and if any one of the calculated statistical feature values is judged to be "normal state", it is determined as <leakage judgment suspended>. to do;
When calculating the statistical feature values in the signal analysis control device 3, frequency skewness, frequency center and regularity are calculated using the spectrum of the acoustic emission signal measured by the inlet AE sensor 1a. Calculate three statistical feature values of the Normal negative log-likelihood; 3 of normal negative log-likelihood, variance frequency and standard deviation using the spectrum of the acoustic emission signal at the outlet measured by the outlet AE sensor 1b calculate branch statistical feature values; Three statistical characteristics of standard deviation, root variation frequency, and frequency RS overall using the vibration signal measured by the inlet vibration sensor 2a and its spectrum calculate a value; Using the spectrum of the vibration signal at the outlet measured by the outlet vibration sensor 2b, frequency skewness, frequency RMS, and normal negative log-likelihood ) A leak detection method for a fluid transport pipe, characterized in that for calculating three statistical feature values. delete

Images