JPH11201859A - Method for detecting leak in pipe by frequency band division - Google Patents

Abstract

PROBLEM TO BE SOLVED: To provide a detecting method in which a water leaking place can be detected simply only by a detecting operation by using microphones by a method in which the detection signal of a pipe flow sound fetched by the microphones at both ends of an inspection section is divided into a plurality of frequency bands and the mean of their correlation coefficients is found. SOLUTION: Underwater microphones 12A, 12B which are installed at both ends of a water-leaking inspection section in a water supply pipe 10 detect the propagation sound of water inside the water supply pipe 10. A controller collects the sound, and it computes a water leaking place together with a sound velocity. At this time, a water leaking sound, in a water leaking point P, which is detected in the installation (reference) point A of the underwater microphone 12A and that which is detected in the installation (comparison) point B of the underwater microphone 12B have the same waveform. Since their distances are different, a delay time τm is generated, and a distance up to the water leaking point P from the reference point A can be found. At this time, underwater sound detection signals are passed through frequency bands in which filter banks 16A, 16B are different, they are input to a correlation-coefficient computing part, and their correlation coefficients are computed for every detection signal whose frequency band is divided. They are added and average by a correlation-coefficient-average processing part 20, and the delay time τm is found. A sound velocity can be computed on the basis of the delay time τm .

Description

DETAILED DESCRIPTION OF THE INVENTION

[0001]

TECHNICAL FIELD The present invention relates to clean water, LP gas,
The present invention relates to a leak detection method in a piping system of city gas, crude oil, a chemical substance (chemical plant), and the like, and more particularly to a pipe leak detection method by frequency band division effective for detecting a leak point in a water network.

[0002]

2. Description of the Related Art It is extremely important to prevent leakage of water, LP gas, city gas, crude oil, chemical substances (chemical plants) and the like in a fluid piping network in order to effectively utilize resources. For this reason, various leak inspections have been conventionally performed on piping networks. For example, the conventional general method of water leakage inspection for a water supply network is called a sound hearing method. Detecting or placing a vibration sensor on the ground along the water pipe to detect the presence or absence of water leakage based on the vibration sound transmitted from the water pipe.

However, in the above-described sound hearing method, vibrations and the like caused by the traffic of a vehicle affect the measurement, and the work must be performed at night. Had the disadvantage of being extremely short. In addition, since listening to sound requires skill, there is also a problem that few workers are engaged in water leak detection.

[0004]

SUMMARY OF THE INVENTION From such a viewpoint,
A method in which a microphone is inserted into a pipe and detected while moving along the pipe (Japanese Patent Application Laid-Open Nos. 50-118554 and 56-160500), and a differential pressure is detected while a pipe flow rate is reduced by an orifice. However, methods for detecting the amount of water leakage have been proposed.

However, in any of the above-mentioned conventional methods, it is necessary to perform an operation such as draining water from a water distribution pipe or shutting off flowing water at the time of actual measurement, and to accurately detect a leakage point while maintaining a steady flow state. I couldn't do that.

The present invention is directed to a method for detecting leakage of a fluid pipe in a flowing state, focusing on the above-mentioned conventional problems,
It is an object of the present invention to provide a leak detection method that can easily detect a leak point only by an operation of detecting a flowing sound by a microphone at both ends of an inspection section of a fluid pipe.

[0007]

In order to achieve the above object, a method for detecting leakage of a pipe by frequency band division according to the present invention is a method for detecting leakage of a fluid pipe in a flow-through state, comprising: Detects fluid sound through microphones at both ends of the filter, passes a detection signal captured by each microphone through a band-pass filter, divides the detection signal into a plurality of frequency bands, and repeats a process of calculating a cross-correlation coefficient by the number of filters. Then, the average of the cross-correlation coefficients of each frequency band is obtained to specify the leak location of the pipe.

[0008]

DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Hereinafter, a specific embodiment of a pipe leak detecting method based on frequency band division according to the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram of an apparatus for applying the method for detecting a pipe leak by frequency band division according to the present invention to a leak test of a water supply pipe network. As shown in the drawing, a water leakage inspection section is set for a water distribution pipe 10 to be inspected, and underwater microphones 12 (12A, 12B) are installed at both ends. The underwater microphone 12 is attached to a T-tube provided at an appropriate interval in the water distribution pipe 10 and
An amplifier for amplifying the sound detected in step 2, and the amplified signal is sent to a control management center using a communication line device or the like. Therefore, the sound propagating through the water in the water distribution pipe 10 is detected by the underwater microphone 12 and collected by the controller in the management center and stored and stored. The section between the pair of sensors including the underwater microphone 12 is a detection section, and the detection section is updated and moved by the movement of the microphone 12. The controller calculates the sound velocity and calculates the leak location as described later, but the control management center simultaneously displays the water supply piping network of the monitoring area, the leak detection section, and the calculation result of the leak amount. Or a video projector.

The operation of the controller performs the following processing. The detection signals of the underwater sound are output from the microphones 12A and 12B to the A / D converters 14 (14A and 14A).
B) to convert it into a digital signal. A /
A filter bank 16 (16A, 16B) that divides a filtering band into a plurality of parts in order to discriminate the detection signal into a plurality of different frequency bands is connected to an output terminal of the D converter 14, and the detection signal is divided for each divided band. , And the signals in the bands that have passed through both filter banks 16A and 16B are input to the cross-correlation coefficient calculator 18 to calculate the signal similarity.

The filter bank 16 divides the detection range frequency (for example, 300 Hz to 1300 Hz) into 10 units of 100 Hz, and 300 to 400 Hz, 400 to 500
Hz,..., And a plurality of band-pass filter groups that pass the respective band ranges of 1200 to 1300 Hz.
The cross-correlation coefficient calculator 18 calculates the cross-correlation coefficient r _xy (m) for each band-divided detection signal based on the following equation.

(Equation 1)

Here, as shown in FIG. 1, in a certain water leakage inspection section, underwater transmission sounds are detected by a pair of underwater microphones 12 located at both ends of the section. Assuming that the installation point of one underwater microphone 12A is a reference point A, the installation point of the other underwater microphone 12B is a control point B, and the water leak point is P, the water leak sound at the water leak point P detected at the reference point A and the control point The water leak sound at the water leak point P detected at B has the same or very similar waveform, but a difference in propagation time occurs because the distance from P to A or B is different. By knowing this propagation time difference, the distance la from the reference point A to the water leak point P can be obtained by the following equation.

(Equation 2)

Where L is the distance between points A and B, and τ
m is the water leak sound propagation time difference, and C is the water leak sound propagation speed. Therefore, if the difference in the water leak sound propagation time is known, the position of the water leak point P can be specified.

The underwater microphones 12 (12A, 12B) simultaneously detect signals from a plurality of sound sources, and the detected sound is captured as a synthesized sound over multiple frequency ranges. The detection sound including the water leak sound reaches the underwater microphones 12A and 12B at both ends with a time difference corresponding to the distance.

Therefore, the sound detected by the microphone 12 is filtered by the above-described filter bank 16 and the waveform of the sound in the water pipe detected by one microphone 12A is used as an original waveform for each divided band, and detected by the other microphone 12B. By calculating the cross-correlation coefficient of the waveform of the water distribution pipe sound as a filtered waveform, a correlation coefficient result for each band frequency as shown in FIG. 2 is obtained. Next, an average correlation coefficient is calculated by adding the correlation coefficient over the entire band and dividing by the number of bands, and this is graphed.

Therefore, the cross-correlation coefficient calculator 18
On the output side is a cross-correlation coefficient averaging unit 2 for each frequency band.
0 is provided. Here, a graph as shown in FIG. 3 is obtained by performing an averaging process for each of the divided bands.
As can be understood from this graph, the time until the peak value of the signal is τm, and this τm is the delay time. By substituting this into the following equation together with the known distance L between microphones, the sound velocity can be easily calculated.

(Equation 3) Here, L is the distance between the paired sensors 12.

Thereafter, the distance la from the reference point A to the water leak point P can be obtained by performing calculations based on the two equations, and the leak location can be specified.

[0018]

As described above, the pipe leak detection method based on frequency band division according to the present invention is a method for detecting leak of a fluid pipe in a flowing state, and the microphone is provided at both ends of a test section through microphones. Detecting a pipe flowing sound, passing a detection signal captured by each microphone through a band-pass filter, dividing the signal into a plurality of frequency bands, and repeating a process of calculating a cross-correlation coefficient by the number of filters, each frequency band The cross-correlation coefficient is calculated to determine the leak location of the pipe, so that the leak location can be easily detected only by detecting the sound with the underwater microphones at both ends of the inspection section of the fluid pipe. The effect is obtained.

[Brief description of the drawings]

FIG. 1 is a configuration block diagram for realizing a pipe leak detection method based on frequency band division according to an embodiment.

FIG. 2 is a graph showing a cross-correlation coefficient of a detection signal for each band obtained by band division.

FIG. 3 is a graph obtained by adding values equal to or greater than a threshold value of a cross-correlation coefficient.

[Explanation of symbols]

 Reference Signs List 10 water distribution pipe 12 (12A, 12B) underwater microphone 14 (14A, 14B) A / D converter 16 (16A, 16B) filter bank 18 cross-correlation coefficient calculation unit 20 cross-correlation coefficient average processing unit

Claims (1)

[Claims] 1. A method for detecting leakage of a fluid pipe in a flowing state, wherein the pipe flowing noise is detected through microphones at both ends of an inspection section, and a detection signal captured by each microphone is subjected to a band-pass filter. To divide into a plurality of frequency bands, repeat the process of calculating the cross-correlation coefficient by the number of filters, determine the average of the cross-correlation coefficient of each frequency band, and specify the pipe leakage point A pipe leak detection method using frequency band division.