Machine Learning Device for Exhaust Gas H2 and O2 Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional exhaust gas analysis devices require separate analyzers for measuring H2 and O2 concentrations, which increases device size and complexity, as these gases do not absorb infrared rays, limiting the capability of FTIR analyzers.
Innovation Solution
A machine learning device that irradiates combustion exhaust gas with light, performs spectral analysis, and uses machine learning to calculate H2 or O2 concentrations based on reference values and spectrum data, enabling measurement without dedicated analyzers by generating correlation data from training data including reference values, spectrum data, and individual component concentrations.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Adaptability or versatility
If a FTIR analyzer is used to analyze exhaust gas components, then multiple components such as CO, CO2, NO, H2O, NO2, C2H5OH, HCHO, or CH4 can be simultaneously analyzed, but components that do not absorb infrared rays such as H2 and O2 cannot be analyzed
Solution Approach 1:
The patent combines the FTIR analyzer with a machine learning unit to create an integrated system. The machine learning unit processes spectrum data from the FTIR analyzer and calculates concentrations of H2 and O2 using trained correlation data, merging the capabilities of infrared spectroscopy with computational analysis to detect all exhaust components through a single device.
Solution Approach 2:
The patent introduces machine learning correlation data as an intermediary between the FTIR spectrum data and the H2/O2 concentration values. This correlation data, trained using reference values from dedicated analyzers, acts as a mediator that translates infrared spectrum information into accurate H2 and O2 concentration measurements without requiring direct infrared absorption from these gases.
2Measurement precision
If dedicated H2 analyzer or O2 analyzer is added to measure H2 or O2 concentration, then measurement capability is improved, but installation space is increased and device size is increased
Solution Approach 1:
The patent makes the FTIR analyzer multi-functional by enabling it to measure not only components that absorb infrared rays but also H2 and O2 that do not absorb infrared rays. The machine learning unit processes the same spectrum data to derive concentrations of all components including H2 and O2, allowing a single device to perform functions previously requiring multiple separate analyzers.
Solution Approach 2:
The patent creates a virtual copy of the dedicated analyzer's measurement capability through machine learning. Instead of physically adding another analyzer, the system uses trained correlation data to replicate the H2 and O2 measurement function computationally, achieving the same measurement capability without the physical hardware overhead.
3Measurement precision
If dedicated H2 analyzer or O2 analyzer is added to measure H2 or O2 concentration, then measurement capability is improved, but device complexity is increased
Solution Approach 1:
The patent replaces the mechanical/physical system of dedicated H2 and O2 analyzers with a computational system. The machine learning unit uses algorithms and trained correlation data to calculate concentrations, substituting complex physical measurement mechanisms with information processing that leverages the existing FTIR spectrum data.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables accurate measurement of H2 and O2 concentrations within the exhaust gas analysis device, improving measurement accuracy and reducing device size by utilizing machine learning to correlate spectral data with component concentrations, even for gases that do not absorb infrared light.
Implementation Method 1
an FTIR analyzer using Fourier transform infrared spectroscopy (FTIR) is used to analyze components contained in exhaust gas. With this FTIR analyzer, it is possible to simultaneously analyze multiple components such as CO, CO2, NO, H2O, NO2, C2H5OH, HCHO, or CH4 in the exhaust gas.
Data Source
AI summary
A machine learning device used in an exhaust gas analysis device that irradiates combustion exhaust gas with light, performs a detection of light transmitted through the combustion exhaust gas, and analyzes the combustion exhaust gas based on a detection signal includes a training data reception unit that receives training data including a reference value of a specific component concentration and at least one of spectrum data obtained by irradiating the combustion exhaust gas with light or an individual component concentration selected based on an element balance formula for determining the specific component concentration, or an arithmetic value of a specific component concentration calculated using the individual component concentration in the element balance formula, and a machine learning unit that performs machine learning on a relationship between the reference value and at least one of the spectrum data, the individual component concentration, or the arithmetic value using the training data.