Chemical Reaction Simulation With Reduced Reaction Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for simulating complex chemical reaction networks face challenges in universality across various process conditions and chemical reaction systems, leading to inefficiencies and inaccuracies due to high computational complexity and insufficient consideration of critical reactions, which limits the automation and simplification of these networks.
Innovation Solution
A processor-implemented method performs a low-dimensional simulation to generate a reduced chemical reaction database by excluding non-essential reactions and species, constructing a graph-based model with importance scoring, and performing target simulations using this reduced database to maintain accuracy and efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Loss of information
If all chemical reaction information is directly reflected in graph modeling, then completeness of chemical reaction data is improved, but computational complexity increases and simulation efficiency deteriorates
Solution Approach 1:
The patent segments the chemical reaction network into two groups: key chemical reactions that significantly affect simulation results and non-key chemical reactions that can be excluded. This segmentation is achieved through sensitivity analysis that identifies reactions with high impact on the system behavior, allowing the model to focus computational resources on critical reactions while omitting less important ones, thus reducing computational complexity while maintaining essential information completeness.
Solution Approach 2:
The patent extracts and removes non-key chemical reactions from the simulation model based on sensitivity analysis results. By identifying and eliminating reactions that have minimal impact on simulation outcomes, the method reduces the number of reactions that need to be processed, thereby decreasing computational complexity and improving simulation efficiency without significantly compromising the completeness of chemically relevant information.
2Productivity
If expert screening is used to simplify chemical reaction networks, then simulation efficiency is improved, but universality across various process conditions deteriorates
Solution Approach 1:
The patent implements a dynamic approach where the identification of key chemical reactions is not fixed but adapts to different process conditions. Through sensitivity analysis, the method dynamically determines which reactions are critical under specific conditions (such as different temperatures, pressures, or feed compositions), allowing the simplified model to maintain both high simulation efficiency and universality across various operating scenarios.
Solution Approach 2:
The patent changes the parameters used for identifying key reactions based on process conditions. Instead of using a static expert-defined list, the method adjusts the sensitivity thresholds and selection criteria according to specific process parameters, enabling the same framework to efficiently simplify chemical reaction networks across diverse conditions while maintaining accuracy.
3Adaptability or versatility
If graph network methods are used without prior knowledge, then adaptability to various systems is improved, but accuracy in identifying critical reactions deteriorates
Solution Approach 1:
The patent incorporates feedback mechanisms where sensitivity analysis results are used to iteratively refine the identification of key chemical reactions. The graph network method processes chemical reaction data and generates sensitivity metrics, which then feed back into the reaction selection process to improve the accuracy of identifying critical reactions. This feedback loop enables the adaptable graph-based approach to progressively enhance its accuracy without requiring prior expert knowledge.
Data Source
Figure 1
Figure 2
Figure 3A
AI summary
A processor-implemented method includes performing, based on an original chemical species database (DB) and an original chemical reaction DB, a low-dimensional simulation based on lower dimensional data and/or determinations than those of a target simulation corresponding to a chemical reaction to be analyzed, the original chemical species DB comprising information corresponding to a chemical species used for the target simulation, and the original chemical reaction DB comprising information corresponding to a chemical reaction used for the target simulation, generating a reduced chemical reaction DB by excluding information corresponding to one or more chemical reactions from the original chemical reaction DB based on a result of the low-dimensional simulation, and performing the target simulation based on the reduced chemical reaction DB and a final chemical species DB.