H2O2 Concentration Detection via Machine Learning Evaluation
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for determining H2O2 concentration in a sterilization process of production machines rely on physical sensors, which can be inefficient and require additional equipment, leading to potential inaccuracies and increased costs.
Innovation Solution
A method using machine learning to evaluate environmental and process-specific parameters, such as ambient air conditions and evaporator device settings, to indirectly determine H2O2 concentration without the need for physical sensors, leveraging existing monitoring device sensors and creating an evaluation model for reliable substance concentration detection.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If a physical sensor is used to detect H2O2 concentration directly, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The patent uses machine learning models as an intermediary to infer H2O2 concentration from environmental parameters (temperature, humidity, pressure) rather than direct sensor measurement. The model acts as a mediator that translates readily available process data into concentration estimates, avoiding the need for complex physical sensors while maintaining measurement capability
Solution Approach 2:
The patent replaces the mechanical/physical sensor system with an information-processing system based on machine learning. Instead of using physical sensors to detect H2O2 molecules, the system uses algorithms to analyze correlations between environmental parameters and concentration, substituting a computational approach for a physical measurement approach
2Measurement precision
If a physical sensor is installed in the piping system, then substance concentration can be detected, but the sensor consumes H2O2 and allows vapor escape, reducing process efficiency
Solution Approach 1:
The machine learning model serves as an indirect measurement intermediary that does not physically interact with the H2O2 vapor. By inferring concentration from environmental parameters measured by non-intrusive sensors, the system avoids the H2O2 consumption and vapor escape problems associated with physical sensors installed in the piping system
Solution Approach 2:
The system uses existing environmental sensors and process data that are already being collected for other purposes to infer H2O2 concentration. This self-service approach leverages available information without requiring additional H2O2-consuming measurement devices, thereby maintaining process efficiency
3Measurement precision
If multiple sensors are added to monitor various parameters, then measurement precision is improved, but device complexity and cost increase
Solution Approach 1:
The machine learning model performs multiple functions by analyzing various environmental parameters (temperature, humidity, pressure) simultaneously to infer H2O2 concentration. This multi-functional approach allows the system to monitor multiple process aspects using a single integrated model rather than requiring separate dedicated sensors for each parameter
Solution Approach 2:
The patent combines multiple environmental parameter measurements and process data into a unified machine learning model that outputs H2O2 concentration estimates. By merging these different data streams and analysis functions into a single system, the patent reduces overall device complexity compared to using multiple separate sensors and analysis systems
Data Source
Figure 1~2
Figure 3
AI summary
The invention proposes a method for determining a substance concentration, in particular an H2O2 concentration, in a gas mixture in a sterilization process of a production machine, wherein, in at least one method step (14), at least one environment-specific and/or process-specific parameter that differs from the substance concentration is captured by means of a sensor (16, 18, 20, 22, 24, 26, 28, 30, 32, 34, 36, 38, 40, 42, 44, 46, 48) of a monitoring apparatus of the production machine, and wherein, in at least one method step (52), the at least one captured environment-specific and/or process-specific parameter is supplied to an evaluation model, created by means of a machine learning process, for evaluation, and the substance concentration, in particular the H2O2 concentration, in the gas mixture is deduced on the basis of the evaluation.