DE102022122295A1 - Method and sensor device for acoustic monitoring of a measuring point on a fitting through which fluid flows and the relevant sensor device - Google Patents

Abstract

The invention relates to a method (100) for acoustically monitoring a measuring point (6) on a fitting (2) through which fluid flows, in particular on a condensate drain (4). According to the invention, the method (100) has the steps: detecting (102) an ambient noise Sound emission (Su) in an area surrounding the measuring point (6); Detecting (104) a structure-borne sound emission (SK) which is emitted by the measuring point (6); Determining (106a) a first frequency spectrum (16a) of the ambient noise sound emission (Su) for a first frequency range (26a); Determining (106b) at least one further frequency spectrum (16b) of the ambient noise sound emission (Su) for at least one further frequency range (26b); Determining (108a) a first frequency spectrum (18a) of the structure-borne sound emission (SK) for the first frequency range (26a); Determining (108b) at least one further frequency spectrum (18b) of the structure-borne sound emission (SK) for the at least one further frequency range (26b); Determining (114a) a first characteristic number (K1) from the frequency spectra (16a, 18a) in the first frequency range (26a); Determining (114b) a further characteristic number (K2) from the further frequency spectra (16b, 18b) in the further frequency range (26b), the characteristic numbers (K1, K2) forming a characteristic pattern (20), determining (116) one Operating state (B) of the steam trap (4) based on the characteristic pattern (20).

Description

The invention relates to a method for acoustically monitoring a measuring point on a fitting through which fluid flows, in particular on a steam trap. The invention further relates to a sensor device for acoustic monitoring of a measuring point on a fitting through which fluid flows, a computer program and a computer-readable medium.

Methods for acoustically monitoring a measuring point on a fitting through which fluid flows, in particular on a steam trap, are known from the prior art. Measuring devices are known for this purpose, which are placed directly on a steam trap using a measuring tip. Such a measuring device measures the intensity of the structure-borne noise generated by the steam trap, especially in the ultrasonic frequency range, and derives from this the set operating state of the steam trap being examined. When it comes to steam traps, it is very important to detect malfunctions at an early stage. Such a malfunction is, for example, an unwanted loss of steam, which is accompanied by specific acoustic features, which are detected in particular based on the emitted structure-borne noise. Although the contact-based measuring instruments known from the prior art have proven themselves, there is still room for improvement.

Since steam traps are often installed in complex steam-conducting systems, it occasionally proves to be challenging to establish direct contact with the steam trap and to place the measuring instruments known from the prior art on the steam trap for the purpose of measurement. This may pose a potential risk to the operator in question, who may have to come very close to a steam trap that may be very hot.

Against this background, the invention was based on the object of developing a method for acoustic monitoring of a measuring point on a fitting through which fluid flows or a corresponding sensor device in such a way that the disadvantages found in the prior art are eliminated as far as possible. In particular, a method and a sensor device are to be specified which increase the reading comfort for the operator and can also be used safely in places that are difficult to access and in very hot steam traps.

According to the invention, the object is achieved in the method of the type mentioned at the outset by the steps: detecting an ambient sound emission in an area surrounding the measuring point, detecting a structure-borne sound emission which is emitted by the measuring point, determining a first frequency spectrum of the ambient sound emission for a first Frequency range, determining at least one further frequency spectrum of the ambient noise sound emission for at least one further frequency range, determining a first frequency spectrum of the structure-borne sound emission for the first frequency range, determining at least one further frequency spectrum of the structure-borne sound emission for the at least one further frequency range, determining a first characteristic Key figure from the frequency spectra in the first frequency range, determining a further characteristic key figure from the further frequency spectra in the further frequency range, the characteristic key figures forming a characteristic pattern, determining an operating state of the steam trap based on the characteristic pattern.

The invention makes use of the knowledge that, based on the method steps mentioned, an operating state of the valve or the steam trap can be determined independently of the need for a contact-based structure-borne sound measurement. This is possible in particular because ambient noise from the typically relatively weak structure-borne sound signal of a steam trap is recorded and taken into account in a corresponding evaluation. In this way, a characteristic pattern can be determined that serves as an indicator of the set operating state of the steam trap.

In the present case, a characteristic pattern is understood to mean a pattern which is formed from individual key figures, each of which is derived from a frequency spectrum, the characteristic pattern allowing conclusions to be drawn about the operating state of the valve. The frequency ranges preferably have no overlap with respect to the selected frequencies and are optionally arranged adjacent to one another. The frequency ranges preferably lie in a total frequency range of 0 kHz to 100 kHz. The frequency ranges preferably have a range width of 1 kHz to 30 kHz, in particular 10 kHz to 25 kHz.

According to a preferred embodiment, the detection of the ambient noise emission and the detection of the structure-borne noise emission are carried out without contact. A non-contact measurement means that measurement in which there is no direct contact between the Measuring equipment, such as a measuring tip, and the fitting or the steam trap exists or must exist. Carrying out non-contact measurements increases the flexibility of the measurement data acquisition for the operator and at the same time prevents the operator from having to step into the immediate vicinity of the steam trap, which is preferable in terms of accessibility and health protection due to high temperatures.

According to a preferred embodiment, the ambient noise emission is recorded at a first distance from the measuring point and the structure-borne sound emission is recorded at a second distance from the measuring point, the second distance being smaller than the first distance. The first distance to the measuring point is preferably 10 cm to 30 cm, in particular 20 cm. The second distance to the measuring point is 1 cm to 10 cm, in particular 5 cm. In other words, a measurement of the ambient sound emission, also referred to as a profile measurement, is preferably carried out first to detect ambient noise at a distance of in particular 20 cm from the fitting or the steam trap. The actual useful signal measurement is then carried out in the form of a structure-borne sound emission, preferably in a second step, at a distance of in particular 5 cm from the steam trap. Frequency spectra are then determined from the recorded measurement signals and a pattern characteristic of an operating state of the valve is determined.

According to a preferred embodiment, the method further comprises the steps: determining a first difference function from the first frequency spectrum of the ambient noise acoustic emission and from the first frequency spectrum of the structure-borne noise emission, determining at least one further difference function from the further frequency spectrum of the ambient noise acoustic emission and the further frequency spectrum the structure-borne sound emission, determining the first characteristic number based on the first difference function, determining the further characteristic number based on the further difference function.

The invention is further developed in that the method has the steps: determining a first integral of the first difference function, determining at least one further integral of the further difference function, determining the first characteristic number based on the first integral, determining the further characteristic number based on the further Integrals. In this way, the areas formed by the difference functions of the frequency spectrums are used in particular for the frequency ranges. This has proven to be particularly suitable for determining the characteristic pattern to identify the operating status of the valve or steam trap.

Preferably, determining an operating state of the steam trap based on the characteristic pattern includes at least one of the following operating states: normal operation of the steam trap, defect in the steam trap.

The method is further developed by the step: Quantitatively determining a derived amount of steam loss and/or condensate based on the characteristic pattern. In this way, the operating state of the steam trap is determined based on the characteristic pattern and/or the derived amount of steam loss and/or condensate is estimated based on the characteristic pattern. The amount of condensate is preferably determined in a frequency range from 0 kHz to 20 kHz. The amount of steam loss is preferably determined in a frequency range from 40 kHz to 70 kHz.

The method is further developed by the step: Providing or detecting a temperature of the steam trap, in particular at the measuring point, with the determination of the operating state and/or the amount of steam loss and/or condensate derived based on the characteristic pattern and the temperature. The additional consideration of the temperature when determining the operating state or the amount of steam loss and/or condensate has proven to be suitable for improving the accuracy in determining the operating state or the derived amount of steam loss and/or condensate.

The method is further developed in that the operating state and the derived amount of steam loss and/or condensate are determined from the characteristic pattern on the basis of machine learning, in particular using pattern recognition. Preferably, a neural network is trained with training data which relates the characteristic pattern to an operating state or a derived steam loss and/or condensate quantity of a steam trap of a specific design. After training the neural network, it can be used to determine the respective operating state or the derived amount of steam loss and/or condensate from the characteristic pattern for a corresponding type of steam trap or a group of steam traps.

The invention has been described above with reference to a method. In a further aspect, the invention relates to a sensor device for acoustically monitoring a measuring point on a fitting through which fluid flows, in particular on a steam trap. The invention solves the problem described at the outset in relation to the sensor device in that it has a sound sensor which is set up to sense an ambient noise emission and a structure-borne sound emission which is emitted by the measuring point in a non-contact manner, and one with the Sound sensor data-conducting control device connected, the control device being set up to carry out the method according to one of the above exemplary embodiments. The sensor device makes use of the same advantages and preferred embodiments as the method according to the invention. In this regard, reference is made to the above statements and their content is included here.

According to a preferred embodiment, the sound sensor is designed as a broadband microphone, in particular as an ultrasonic microphone. Since the characteristic sound frequencies, which allow conclusions to be drawn about the operating state of the steam trap, are particularly in the ultrasonic range, the sound sensor is preferably designed as an ultrasonic microphone. According to a preferred embodiment, the sensor device is designed as a mobile device. In this way, the device can be carried particularly easily by the operator and can also be used in confined spaces.

According to a preferred embodiment, the sensor device has a display means, in particular a display, which is set up to display the operating state of the fitting and/or the amount of steam loss and/or condensate. In this way, the operator receives the desired information about the condition of the valve or steam trap directly on site.

The invention is further developed in that the sensor device has a light source for illuminating the measuring point or the steam trap. In this way, the visibility and correct positioning of the sensor device at the measuring point of the condensate drain is made easier in difficult-to-access and dark places. According to a preferred embodiment, the sensor device has a distance meter, in particular a laser distance meter. In this way, the correct spacing of the sensor device from the measuring point for carrying out the ambient noise emission or the structure-borne sound emission can be facilitated and monitored.

According to a preferred embodiment, the steam trap or the measuring point has a marking that makes it easier to carry out the measurement correctly. The marking is preferably designed as a QR code or barcode, with the sensor device having a corresponding scanner. In this way, the type of steam trap can be directly linked to the measured values collected. In addition, the determination of the characteristic pattern or the operating state can be directly adapted to the type of steam trap detected. According to an alternative embodiment, near field communication (NFC) is used to identify the steam trap.

The sensor device is further developed in that it has a temperature sensor, in particular an infrared thermometer, wherein the temperature sensor is set up to sense a temperature of the steam trap, in particular to sense it in a non-contact manner, and wherein the temperature sensor is connected to the control device in a data-conducting manner. The use of a temperature sensor, which also enables non-contact measurement, has proven, as stated, to be preferred for increasing the measurement accuracy or the determination accuracy of the operating state of the fitting and the amount of steam loss and/or condensate.

Preferably, the control device is connected to a communication interface in a data-conducting manner. The communication interface is set up in particular to communicate wirelessly with a cloud, a mobile device or an external network or computer via a wireless network. In this way, the measurement data can be transmitted to a corresponding system in real time or with a time delay and is therefore available and visible in a central system control and monitoring system, for example.

In a further aspect, the invention relates to a computer program comprising commands which cause a sensor device designed according to one of the above exemplary embodiments to carry out the method according to one of the above exemplary embodiments. In a further aspect, the invention relates to a computer-readable medium on which the computer program according to the above exemplary embodiment is stored. The computer program and the computer-readable medium make use of the same advantages and preferred embodiments as the method according to the invention and the sensor device according to the invention and vice versa. In this regard, reference is made to the above statements and their content is included here.

• 1 ein Blockschaltbild eines erfindungsgemäßen Verfahrens;
• 2 eine schematische Darstellung einer erfindungsgemäßen Sensoreinrichtung in einer schematischen Darstellung;
• 3a beispielhafte Frequenzspektren einer Umgebungsgeräusch-Schallemission und einer Körperschall-Emission;
• 3b eine aus den Frequenzspektren der 3a gebildete Differenzfunktion sowie ein charakteristisches Muster der Differenzfunktion
• 4 ein Ausführungsbeispiel eines erfindungsgemäßen Computerprogramms in einer schematischen Darstellung; und
• 5 eine schematische Darstellung eines erfindungsgemäßen computerlesbaren Mediums.

The invention is described in more detail below and using a preferred exemplary embodiment with reference to the attached figures. Show here:

• 1 a block diagram of a method according to the invention;
• 2 a schematic representation of a sensor device according to the invention in a schematic representation;
• 3a exemplary frequency spectra of an ambient sound emission and a structure-borne sound emission;
• 3b one from the frequency spectra of the 3a formed difference function as well as a characteristic pattern of the difference function
• 4 an exemplary embodiment of a computer program according to the invention in a schematic representation; and
• 5 a schematic representation of a computer-readable medium according to the invention.

1 shows a block diagram of a method 100 for acoustic monitoring of a measuring point 6 on a fitting 2 through which fluid flows, in particular on a steam trap 4, which in 2 is shown. The method 100 has the steps: Detecting 102 an ambient noise acoustic emission Su in an environment in 2 measuring point 6 shown, detecting 104 a structure-borne sound emission S _K , which is emitted by the measuring point 6, determining 106a a first frequency spectrum 16a of the ambient noise sound emission Su for a first frequency range 26a, which is exemplified in 3a is shown. The method 100 further has the steps: determining 106b at least one further frequency spectrum 16b of the ambient noise sound emission Su for at least one further frequency range 26b, determining 108a a first frequency spectrum 18a of the structure-borne sound emission S _K for the first frequency range 26a and determining 108b at least one further frequency spectrum 18b of the structure-borne noise emission S _K for the at least one further frequency range 26b.

In method step 110a, a first difference function 22a is determined from the first frequency spectrum 16a of the ambient noise emission Su and from the first frequency spectrum 18a of the structure-borne sound emission S _K. Furthermore, at least one further difference function 22b is determined 110b from the further frequency spectrum 16b of the ambient noise emission Su and the further frequency spectrum 18b of the structure-borne sound emission S _K. In method step 112a, a first integral I1 is then determined from the first difference function 22a and in method step 112b at least one further integral I2 of the further difference function 22b.

Then, in method steps 114a and 114b, a first characteristic number K1 is determined based on the first integral I1 and a further characteristic number K2 is determined based on the further integral I2. Then, in method step 116, an operating state B of the steam trap 4 is determined based on the characteristic pattern 20, which is formed by the characteristic key numbers K1 and K2. Finally, in method step 118, a derived amount of steam loss and/or condensate is quantitatively determined based on the characteristic pattern 20.

The operating state B and the derived amount of steam loss and/or condensate are preferably determined based on the characteristic pattern 20 and a temperature T of the steam trap. The determination 116, 118 of the operating state B and the derived amount of steam loss and/or condensate from the characteristic pattern 20 is carried out in particular on the basis of machine learning.

2 shows an exemplary embodiment of a sensor device 1 for acoustic monitoring of a measuring point 6 on a system 2 through which fluid flows, in particular on a condensate drain 4. The sensor device 1 has a sound sensor 8. The sound sensor 8 is set up to sense ambient noise sound emission Su without contact. The sensor device 1 is also set up to sense a structure-borne sound emission S _K , which is emitted from a location 6, in a non-contact manner. In addition, the sensor device 1 has a control device 12 which is connected to the sound sensor 8 in a data-conducting manner, the control device 12 being set up to carry out the method 100 according to 1 to carry out.

The sound sensor 8 is designed as a broadband microphone 10, in particular as an ultrasonic microphone 10. The sensor device 1 is designed as a mobile device. The sensor device 1 also has a display means 14, which is designed as a display 14. The display means 14 is set up to display the operating state B of the fitting 2 and/or the amount of steam loss and/or condensate. The sensor device 1 also has a temperature sensor 28. The temperature sensor 28 is designed as an infrared thermometer 30. The temperature sensor 28 is set up to detect a temperature temperature T of the steam trap 4, in particular to sense it without contact. The temperature sensor 28 is connected to the control device 12 in a data-conducting manner. The measurement of the ambient noise emission Su takes place at a first distance du from the measuring point 6. The measurement of the structure-borne sound emission S _K takes place at a second distance d _K from the measuring point 6. The second distance d _K is smaller than the first distance du. The first distance from you to the measuring point is preferably 20 cm. A second measurement is then carried out at the second distance d _K from the measuring point 6 of in particular 5 cm.

How out 2 can be seen, the sound sensor 8 or the ultrasonic microphone 10 is connected to a filter and amplifier 32 in a data-conducting manner. The filter and amplifier 32 amplifies and filters the sound signal determined by the sound sensor. The filter and amplifier 32 is in turn connected to an analog-digital converter 34. By means of the analog-digital converter 34, the analog signal is converted into a digital signal, which is then fed to the control device 12. The control device 12 is connected to a communication point 36 in a data-conducting manner. The communication interface 36 is set up to communicate with a cloud 38, a mobile device 40 and/or a computer 42 via a data network 44.

3a shows a frequency spectrum 16 of an exemplary ambient noise sound emission Su and a frequency spectrum 18 of a structure-borne sound emission _SK , whereby for the spectra 16, 18 the frequency is plotted against the sound pressure in a frequency range from 0 to 50 kHz. For a first frequency range 26a, a first frequency spectrum 16a of the ambient noise emission Su and a first frequency spectrum 18a of a structure-borne sound emission S _K can be determined. The same can be done for the further frequency range 26b shown here as an example, which has the frequency spectra 16b and 18b. As in 3b shown, a difference function 22a is then formed for the frequency spectra 16a and 18a, the integral I1 of which in turn forms a first characteristic number K1. For the further frequency range 26b shown as an example, a difference function 22b is also formed, the integral I2 of which forms the further characteristic number K2. The key figures K1 and K2 form a characteristic pattern 20. In addition, in 3b further characteristic key numbers K _n are shown, which are formed from integrals I _n and can contribute to the characteristic pattern 20.

4 shows a computer program 200. The computer program 200 includes commands that cause one to 2 trained sensor device 1 according to the method 100 1 executes. 5 shows a computer-readable medium 300. The computer program 200 is on the computer-readable medium 300 4 saved.

Reference symbols

1

Sensor device

2

fitting

4

Steam trap

6

Measuring point

8th

Sound sensor

10

Ultrasonic microphone

12

Control device

14

Display means/display

16a

first frequency spectrum of ambient noise sound emission

16b

further frequency spectrum of ambient noise noise emission

18a

first frequency spectrum of structure-borne noise emissions

18b

further frequency spectrum of structure-borne noise emissions

20

characteristic pattern

22a

first difference function

22b

another difference function

26a

first frequency range

26b

further frequency range

28

Temperature sensor

30

Infrared thermometer

32

Filters and amplifiers

34

Analog-to-digital converter

36

Communication interface

38

Cloud

40

Mobile device

42

computer

44

Data network

100

Proceedings

102

Detecting ambient noise acoustic emission

104

Detecting structure-borne noise emissions

106a

Determining a first frequency spectrum of the ambient noise sound emission

106b

Determining a further frequency spectrum of the ambient noise sound emission

108a

Determining a first frequency spectrum of the structure-borne noise emission

108b

Determination of a further frequency spectrum of structure-borne noise emissions

110a

Determine a first difference function

110b

Determine another difference function

112a

Determine a first integral

112a

Determine another integral

114a

Determining a first characteristic key figure

114b

Determination of another characteristic key figure

116

Determining an operating state

118

Determine a derived amount of steam loss and/or condensate

200

Computer program

300

Computer readable medium

b

Operating status of the valve

you

Distance to the measuring point for measuring the ambient noise emission

dK

Distance to the measuring point for measuring structure-borne noise emissions

K1

first characteristic key figure

K2

another characteristic key figure

Kn

characteristic key figures

I1

first integral

I2

another integral

In

characteristic integrals

Su

Ambient noise sound emission

SK

Structure-borne noise emission

T

Steam trap temperature

Claims (18)

Method (100) for the acoustic monitoring of a measuring point (6) on a fitting (2) through which fluid flows, in particular on a steam trap (4), with the steps: - detecting (102) an ambient sound emission (Su) in an area surrounding the measuring point ( 6); - Detecting (104) a structure-borne noise emission (S _K ), which is emitted by the measuring point (6); - Determining (106a) a first frequency spectrum (16a) of the ambient noise sound emission (Su) for a first frequency range (26a); - Determining (106b) at least one further frequency spectrum (16b) of the ambient noise sound emission (Su) for at least one further frequency range (26b); - Determining (108a) a first frequency spectrum (18a) of the structure-borne noise emission (S _K ) for the first frequency range (26a); - Determining (108b) at least one further frequency spectrum (18b) of the structure-borne noise emission (S _K ) for the at least one further frequency range (26b); - Determining (114a) a first characteristic number (K1) from the frequency spectra (16a, 18a) in the first frequency range (26a); - Determining (114b) a further characteristic number (K2) from the further frequency spectra (16b, 18b) in the further frequency range (26b), the characteristic numbers (K1, K2) forming a characteristic pattern (20), - Determining (116 ) an operating state (B) of the steam trap (4) based on the characteristic pattern (20). Procedure (100) according to Claim 1 , wherein the detection (102) of the ambient noise emission (Su) and the detection (104) of the structure-borne noise emission (S _K ) take place without contact. Procedure (100) according to Claim 2 , wherein the ambient noise emission (Su) is detected at a first distance (du) from the measuring point (6) and wherein the structure-borne sound emission (S _K ) is detected at a second distance (d _K ) from the measuring point (6), where the second distance (d _K ) is less than the first distance (du). Procedure (100) according to Claim 3 , whereby the first distance (du) to the measuring point (6) is 10 cm to 30 cm, in particular 20 cm. Procedure (100) according to Claim 3 or 4 , whereby the second distance (d _K ) to the measuring point (6) is 1 cm to 10 cm, in particular 5 cm. Method (100) according to one of the preceding claims, further comprising the steps: - determining (110a) a first difference function (22a) from the first frequency spectrum (16a) of the ambient sound emission (Su) and from the first frequency spectrum (18a) of the structure-borne noise -Emission (S _K ), - Determining (11 0b) at least one further difference function (22b) from the further frequency spectrum (16b) of the ambient noise emission (Su) and the further frequency spectrum (18b) of the structure-borne sound emission (S _K ), - determining the first characteristic number (K1) based on the first difference function (22a), determining the further characteristic number (K2) based on the further difference function (22b). Procedure (100) according to Claim 6 , further comprising the steps: - determining (112a) a first integral (I1) of the first difference function (22a), - determining (112b) at least one further integral (I2) of the further difference function (22b), - determining the first characteristic number ( K1) based on the first integral (I1), determining the further characteristic number (K2) based on the further integral (I2). Method (100) according to one of the preceding claims, wherein determining (116) an operating state (B) of the steam trap (4) based on the characteristic pattern (20) comprises at least one of the following operating states (B): normal operation of the steam trap, defect of the steam trap. Method (100) according to one of the preceding claims, further comprising the step: - Quantitatively determining (118) a derived amount of steam loss and/or condensate based on the characteristic pattern (20). Procedure (100) according to Claim 9 , further comprising the step: - Providing or detecting a temperature (T) of the steam trap (4), in particular at the measuring point (6), determining the operating state (B) and / or the derived amount of steam loss and / or condensate on the basis the characteristic pattern (20) and the temperature (T). Procedure (100) according to Claim 9 or 10 , wherein the determination (116, 118) of the operating state (B) and the derived amount of steam loss and / or condensate from the characteristic pattern (20) takes place on the basis of machine learning. Sensor device (1) for acoustic monitoring of a measuring point (6) on a fitting (2) through which fluid flows, in particular on a condensate drain (4), with a sound sensor (8), which is set up to detect an ambient sound emission (Su) and a structure-borne sound -Emission (S _K ), which is emitted by the measuring point (6), to be sensed in a contactless manner by a control device (12) connected to the sound sensor (8) in a data-conducting manner, the control device (12) being set up to carry out the method (100). after one of the Claims 1 until 10 to carry out. Sensor device (1). Claim 12 , wherein the sound sensor (8) is designed as a broadband microphone, in particular an ultrasonic microphone (10). Sensor device (1). Claim 12 or 13 , wherein the sensor device (1) is designed as a mobile device (28). Sensor device (1) according to one of Claims 12 - 14 , wherein the sensor device (1) has a display means (14), in particular a display (14), which is set up to display the operating state (B) of the fitting (2) and / or the amount of steam loss and / or condensate. Sensor device (1) according to one of Claims 12 - 15 , further comprising a temperature sensor (28), in particular an infrared thermometer (30), wherein the temperature sensor (28) is set up to sense a temperature (T) of the steam trap (4), in particular to sense it without contact, and wherein the temperature sensor (28) is connected to the control device (12) in a data-conducting manner. Computer program (200), comprising instructions that cause one of the Claims 12 until 16 trained sensor device (1) the method (100) according to one of Claims 1 until 11 executes. Computer-readable medium (300) on which the computer program (200) is written Claim 17 is stored.

Images