Sample-based gas quality control by means of raman spectroscopy

Patent metadata

Field	Value
Publication number	US-11460407-B2
Application number	US-201816636281-A
Country	US
Kind code	B2
Filing date	Aug 2, 2018
Priority date	Aug 4, 2017
Publication date	Oct 4, 2022
Grant date	Oct 4, 2022

Abstract

For the purpose of gas quality monitoring, a spectroscopic examination of a gas sample from a space ( 10 ) to be monitored is carried out, e.g. by Raman spectroscopy. The spectroscopic examination yields a measurement spectrum extending over a wavelength range. A deviation of the measurement spectrum from at least one comparison sample ( 160 ) is then detected. Depending on the detected deviation, a gas quality warning (QW) is produced.

First claim

The invention claimed is:

1. A method for gas quality monitoring, comprising:
spectroscopic examination of a gas sample from a space to be monitored, wherein the spectroscopic examination yields a measurement spectrum extending over a wavelength range;
detecting a deviation of the measurement spectrum from at least one comparison pattern, wherein the measurement spectrum is represented as an image;
depending on the deviation that was detected, producing a gas quality warning; and
depending on the gas quality warning produced, controlling an apparatus for conveying gas into the space to be monitored and/or out of the space to be monitored,
wherein the detecting the deviation of the measurement spectrum comprises comparing the image of the measurement spectrum with the at least one comparison pattern.

2. The method of claim 1 , comprising:
detecting at least one ambient parameter of the space to be monitored; and
selecting the at least one comparison pattern depending on the at least one ambient parameter that was detected.

3. The method of claim 2 ,
wherein the at least one ambient parameter comprises a temperature, a pressure and/or a moisture content.

4. The method of claim 1 , comprising:
detecting the deviation by a comparison of a wavelength-dependent intensity of the measurement spectrum with at least one comparison intensity defined by the at least one comparison pattern.

5. The method of claim 1 , comprising:
presenting the measurement spectrum as the image, which is subdivided into a first region and a second region depending on a wavelength-dependent intensity; and
detecting the deviation by detecting an overlap of the first region and/or of the second region with at least one comparison region defined by the at least one comparison pattern.

6. The method of claim 1 , comprising:
detecting a severity of the deviation; and
producing the gas quality warning depending on the severity of the deviation that was detected.

7. The method of claim 1 ,
wherein the at least one comparison pattern describes shared properties of measurement spectra classified as inconspicuous.

8. The method of claim 1 ,
wherein the spectroscopic examination of the gas sample is carried out by Raman spectroscopy.

9. The method of claim 1 , comprising:
determining the at least one comparison pattern by training a machine-based learning algorithm.

10. The method of claim 1 ,
wherein the apparatus comprises a ventilation installation and the space to be monitored is situated in an air feed section of the ventilation installation or in a space to be ventilated by way of the ventilation installation.

11. The method of claim 1 ,
wherein the apparatus comprises a process installation and the space to be monitored is situated in a feed section for a process gas used in the process installation or is situated in a process space of the process installation.

12. An apparatus for gas quality monitoring, comprising:
a spectrometer device for spectroscopic examination of a gas sample from a space to be monitored, wherein the spectroscopic examination yields a measurement spectrum extending over a wavelength range; and
an evaluation device configured to perform operations comprising:
detecting a deviation of the measurement spectrum from at least one comparison pattern, wherein the measurement spectrum is represented as an image; and
producing a gas quality warning depending on the deviation that was detected; and
depending on the gas quality warning produced, controlling an apparatus for conveying gas into the space to be monitored and/or out of the space to be monitored,
wherein the detecting the deviation of the measurement spectrum comprises comparing the image of the measurement spectrum with the at least one comparison pattern.

13. A ventilation installation, comprising:
an apparatus for conveying air into a space to be ventilated and/or out of the space to be ventilated; and
an apparatus for gas quality monitoring comprising:
a spectrometer device for spectroscopic examination of a gas sample from a space to be monitored, wherein the spectroscopic examination yields a measurement spectrum extending over a wavelength range, wherein the measurement spectrum is represented as an image; and
an evaluation device configured to detect a deviation of the measurement spectrum from at least one comparison pattern based on comparing the image of the measurement spectrum with the at least one comparison pattern, and to produce a gas quality warning depending on the deviation that was detected,
wherein the space to be monitored is situated in an air feed section of the ventilation installation or in the space to be ventilated, and
wherein the apparatus for conveying air is controlled depending on the gas quality warning.

14. A process installation, comprising:
an apparatus for conveying gas into a process space of the process installation and/or out of the process space of the process installation; and
an apparatus for gas quality monitoring comprising:
a spectrometer device for spectroscopic examination of a gas sample from a space to be monitored, wherein the spectroscopic examination yields a measurement spectrum extending over a wavelength range, wherein the measurement spectrum is represented as an image; and
an evaluation device configured to detect a deviation of the measurement spectrum from at least one comparison pattern based on comparing the image of the measurement spectrum with the at least one comparison pattern, and to produce a gas quality warning depending on the deviation that was detected,

wherein the space to be monitored is situated in a feed section for a process gas used in the process installation or is situated in the process space of the process installation, and
wherein the apparatus for conveying gas is controlled depending on the gas quality warning.

Assignees

• Zeiss Carl Jena Gmbh
• Carl Zeiss Spectroscopy Gmbh

Inventors

• Rode Michael
• Gatto Alexandre
• Deparnay Arnaud
• Helgert Michael

Classifications

• G06N3/0464

  Convolutional networks [CNN, ConvNet] · CPC title

• G06N3/09

  Supervised learning · CPC title

• B60H1/008

  the input being air quality · CPC title

• G01N21/65 Primary

  Raman scattering · CPC title

• F24F2110/50

  Air quality properties · CPC title