US-20260126377-A1 - APPARATUSES, METHODS, AND COMPUTER PROGRAM PRODUCTS FOR GAS DETECTION

US20260126377A1US 20260126377 A1US20260126377 A1US 20260126377A1US-20260126377-A1

Abstract

Methods, apparatuses, and computer program products for energy-centric predictive maintenance scheduling are provided. For example, a computer-implemented method may include separately scanning each of a predetermined plurality of different training gases with infrared light at each of a first predetermined plurality of different wavelengths, for each of the predetermined plurality of different training gases, detecting and recording the absorption of the infrared light at each of the different wavelengths, creating a plurality of training absorption waveforms, one training absorption waveform for each possible different combination of each of the predetermined plurality of different training gases at each of a predetermined plurality of different concentrations and at each of a predetermined plurality of different temperatures, determining a plurality of training waveform features of each training absorption waveform, and inputting the plurality of training waveform features for each training absorption waveform into a data model to train the data model.

Inventors

Janmejaya Tripathy
Sumit Suresh KULKARNI
Nimmagadla Lakshmi SNEHITA

Assignees

Honeywell Analytics Inc.

Dates

Publication Date: 20260507
Application Date: 20260102
Priority Date: 20230112

Claims (20)

1 . A computer-implemented method, comprising: separately scanning each of a predetermined plurality of different training gases with infrared light at each of a first predetermined plurality of different wavelengths; recording absorption of the infrared light at each of the first predetermined plurality of different wavelengths; creating a plurality of training absorption waveforms, one training absorption waveform for each possible different combination of each of the predetermined plurality of different training gases at each of a predetermined plurality of different concentrations and at each of a predetermined plurality of different temperatures; inputting features for each training absorption waveform into a data model to train the data model; scanning one or more unknown gases with infrared light at each of a second predetermined plurality of different wavelengths; recording absorption of the infrared light at each of the second predetermined plurality of different wavelengths; inputting features of recorded absorption of the infrared light at each of the second predetermined plurality of different wavelengths into the data model; analyzing the features of the recorded absorption of the infrared light and the features of each training absorption waveform from the data model; and determining identity of the one or more unknown gases based on the analyzing; wherein the one or more unknown gases comprises one or more of the predetermined plurality of different training gases.
2 . The method of claim 1 , wherein the features of at least one of the training absorption waveform and recorded absorption waveform comprises one or more of: a number of peaks, an absorption value at a tallest peak, absorption values at all peaks, a wavelength location of the tallest peak, wavelength locations of all peaks, a full width at half maximum of the peaks, and wavelength zones exhibiting zero absorption.
3 . The method of claim 1 further comprising: creating a detection absorption waveform for the scanned one or more unknown gases; and determining features of the detection absorption waveform.
4 . The method of claim 1 further comprising: generating a concentration of the one or more unknown gases from the data model; and displaying the concentration of the one or more unknown gases on at least one display of a user device.
5 . The method of claim 1 , wherein the second predetermined plurality of different wavelengths equals the first predetermined plurality of different wavelengths or the second predetermined plurality of different wavelengths is a subset of the first predetermined plurality of different wavelengths.
6 . The method of claim 5 , wherein the first and second predetermined plurality of different wavelengths are evenly spaced over a predetermined wavelength range.
7 . The method of claim 1 , wherein separately scanning each of the predetermined plurality of different training gases comprises separately scanning each of the predetermined plurality of different training gases at each of the predetermined plurality of different concentrations.
8 . The method of claim 1 , wherein separately scanning each of the predetermined plurality of different training gases comprises separately scanning each of the predetermined plurality of different training gases at each of the predetermined plurality of different temperatures.
9 . The method of claim 1 further comprising: determining a temperature of the scanned one or more unknown gases; and inputting the determined temperature of the scanned one or more unknown gases into the data model.
10 . The method of claim 1 further comprising determining a lower explosion limit percentage of the scanned one or more unknown gases.
11 . An apparatus comprising at least one processor and at least one non-transitory memory comprising program code, wherein the at least one non-transitory memory and the program code are configured to, with the at least one processor, cause the apparatus to at least: separately scan each of a predetermined plurality of different training gases with infrared light at each of a first predetermined plurality of different wavelengths; record absorption of the infrared light at each of the first predetermined plurality of different wavelengths; create a plurality of training absorption waveforms, one training absorption waveform for each possible different combination of each of the predetermined plurality of different training gases at each of a predetermined plurality of different concentrations and at each of a predetermined plurality of different temperatures; input features for each training absorption waveform into a data model to train the data model; scan one or more unknown gases with infrared light at each of a second predetermined plurality of different wavelengths; record absorption of the infrared light at each of the second predetermined plurality of different wavelengths; input features of recorded absorption of the infrared light at each of the second predetermined plurality of different wavelengths into the data model; analyze the features of the recorded absorption of the infrared light and the features of each training absorption waveform from the data model; and determine identity of the one or more unknown gases based on the analyzes; wherein the one or more unknown gases comprises one or more of the predetermined plurality of different training gases.
12 . The apparatus of claim 11 , wherein the second predetermined plurality of different wavelengths equals the first predetermined plurality of different wavelengths or the second predetermined plurality of different wavelengths is a subset of the first predetermined plurality of different wavelengths.
13 . The apparatus of claim 12 , wherein the first and second predetermined plurality of different wavelengths are evenly spaced over a predetermined wavelength range.
14 . The apparatus of claim 11 , wherein separately scanning each of the predetermined plurality of different training gases comprises separately scanning each of the predetermined plurality of different training gases at each of the predetermined plurality of different concentrations.
15 . The apparatus of claim 11 , wherein separately scanning each of the predetermined plurality of different training gases comprises separately scanning each of the predetermined plurality of different training gases at each of the predetermined plurality of different temperatures.
16 . The apparatus of claim 11 , wherein the at least one non-transitory memory and the program code are further configured to, with the at least one processor, cause the apparatus to at least: determine a temperature of the scanned one or more unknown gases; and input the determined temperature of the scanned one or more unknown gases into the data model.
17 . The apparatus of claim 11 , wherein the at least one non-transitory memory and the program code are further configured to, with the at least one processor, cause the apparatus to at least determine a lower explosion limit percentage of the scanned one or more unknown gases.
18 . A computer program product comprising at least one non-transitory computer-readable storage medium having computer-readable program code portions stored therein, the computer readable program code portions comprising an executable portion configured to: separately scan each of a predetermined plurality of different training gases with infrared light at each of a first predetermined plurality of different wavelengths; record absorption of the infrared light at each of the first predetermined plurality of different wavelengths; create a plurality of training absorption waveforms, one training absorption waveform for each possible different combination of each of the predetermined plurality of different training gases at each of a predetermined plurality of different concentrations and at each of a predetermined plurality of different temperatures; input features for each training absorption waveform into a data model to train the data model; scan one or more unknown gases with infrared light at each of a second predetermined plurality of different wavelengths; record absorption of the infrared light at each of the second predetermined plurality of different wavelengths; input features of recorded absorption of the infrared light at each of the second predetermined plurality of different wavelengths into the data model; analyze the features of the recorded absorption of the infrared light and the features of each training absorption waveform from the data model; and determine identity of the one or more unknown gases based on the analyzing; wherein the one or more unknown gases comprises one or more of the predetermined plurality of different training gases.
19 . The computer program product of claim 18 , wherein the second predetermined plurality of different wavelengths equals the first predetermined plurality of different wavelengths or the second predetermined plurality of different wavelengths is a subset of the first predetermined plurality of different wavelengths.
20 . The computer program product of claim 18 , wherein the computer-readable program code portions further comprise an executable portion configured to: determine a temperature of the scanned one or more unknown gases; and input the determined temperature of the scanned one or more unknown gases into the data model.

Description

CROSS-REFERENCE TO RELATED APPLICATIONS This application is a continuation of and claims priority to pending U.S. patent application Ser. No. 18/405,716, filed Jan. 5, 2024, which in turn claims priority pursuant to 35 U.S.C. 119 (a) to Indian Application No. 202311002476, filed Jan. 12, 2023, which applications are incorporated herein by reference in their entirety. TECHNOLOGICAL FIELD Example embodiments of the present disclosure relate generally to detecting potentially hazardous gases and, more particularly, to methods, apparatuses, and computer program products for providing machine learning and artificial-intelligence-based identification and quantification of potentially hazardous gases. BACKGROUND Many industrial facilities/applications have the potential to produce and/or release one or more gases which may cause a hazardous, sometimes potentially explosive, atmosphere within the facility. Such industrial facilities/applications include, but are not limited to, offshore oil and gas platforms, floating production storage and offloading vessels, tankers, onshore oil and gas terminals, refineries, liquified natural gas bottling plants, gas compressor/metering stations, and gas turbine power plants. Such potentially hazardous gases include, but are not limited to, hydrocarbons such as methane, ethane, propane, and butane. The atmosphere within and around such industrial facilities is typically monitored to detect the presence of such potentially hazardous gases to prevent an accumulation that could result in an explosion. Conventional optical infrared gas detectors are often installed in and around such industrial facilities. Such conventional gas detectors are typically calibrated to detect a single type of gas and are therefore termed “fixed gas detectors.” Such conventional gas detectors provide relatively quick analysis of the atmosphere and detection of the calibrated gas. However, some industrial facilities/applications are capable of producing/releasing multiple different types of hazardous gases. These fixed gas detector are prone to cross sensitivity issues when exposed to other gases in the environment due to cross interference in the spectral absorption properties. Some gases have a stronger absorption peak than the calibrated gas. This can result in a “false alarm” condition, where an alarm is triggered when the cumulative concentration of flammable gas mixture has not reached the predetermined safety limit. More sophisticated gas analyzers, such as those that use Fourier Transform Infrared (FTIR) spectroscopy, are capable of detecting many different gases and combinations of gases due to their ability to scan a large wavelength range with a resolution of about 0.1 nanometer (nm). However, such FTIR gas analyzers are significantly more expensive than conventional single gas detectors and take much longer to complete a scan and detect the gas(es) present, thereby limiting their usability. Moreover, conventional Michelson-type FTIR gas analyzers can be negatively affected by vibration and temperature shift. Applicant has discovered problems with current implementations of gas detection systems and methods. Through applied effort, ingenuity, and innovation, many of these identified problems have been solved by developing solutions that are included in embodiments of the present disclosure, many examples of which are described in detail herein. BRIEF SUMMARY In general, embodiments of the present disclosure provided herein provide improvements in gas detection. Other implementations for gas detection will be, or will become, apparent to one with skill in the art upon examination of the following figures and detailed description. It is intended that all such additional implementations be included within this description be within the scope of the disclosure and be protected by the following claims. In accordance with a first aspect of the disclosure, a method is provided. The method may be computer-executed via one or more computing devices embodied in hardware, software, firmware, and/or a combination thereof, as described herein. An example implementation of the method is performed at a device with one or more processors and one or more memories. The example method includes separately scanning each of a predetermined plurality of different training gases with infrared light at each of a first predetermined plurality of different wavelengths, for each of the predetermined plurality of different training gases, detecting and recording the absorption of the infrared light at each of the first predetermined plurality of different wavelengths, creating a plurality of training absorption waveforms, one training absorption waveform for each possible different combination of each of the predetermined plurality of different training gases at each of a predetermined plurality of different concentrations and at each of a predetermined plurality of different temperatures, determining a plurality of traini