Chemical Reaction Simulation Using Reduced Reaction Networks
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for simplifying complex chemical reaction networks face challenges in universality across various process conditions and chemical reaction systems, leading to inefficiencies and inaccuracies in determining key species and reactions, particularly when computational complexity increases or when graph modeling underestimates critical reactions.
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 with importance scoring, and performing a target simulation 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 the influence of critical chemical reactions may be underestimated and accuracy in determining key species deteriorates
Solution Approach 1:
The patent performs a preliminary low-dimensional simulation before the main graph modeling process to identify and prioritize critical chemical reactions. This preliminary action filters the chemical reaction database to retain only essential reactions, ensuring that when graph modeling is performed, the most influential reactions are properly weighted and not lost among less important ones. This resolves the contradiction by preparing the data in advance to maintain both completeness of important information and accuracy in identifying key species.
Solution Approach 2:
The patent extracts critical chemical reactions from the complete chemical reaction database through low-dimensional simulation and importance analysis. By separating essential reactions from less important ones, the method ensures that graph modeling focuses on the most influential reactions, preventing them from being underestimated while maintaining the overall completeness of the chemical reaction system.
2Measurement precision
If high-dimensional simulations are performed to obtain simplified chemical reaction networks, then accuracy is improved, but computational complexity and time consumption increase significantly
Solution Approach 1:
The patent segments the simulation process into two distinct stages: a low-dimensional simulation phase to identify critical reactions, and a subsequent graph modeling phase using only those essential reactions. This segmentation allows the computationally intensive high-dimensional simulation to be avoided for the entire reaction network, while still achieving accurate results by focusing computational resources only on the most important reactions identified in the first stage.
Solution Approach 2:
Instead of performing complete high-dimensional simulations on all chemical reactions, the patent applies partial action by conducting low-dimensional simulations to identify a subset of critical reactions. This partial approach is sufficient to capture the essential chemistry, avoiding the excessive computational cost of simulating all reactions in high dimension while maintaining accuracy for the most important processes.
3Measurement precision
If expert knowledge is used to pre-screen chemical species and reactions, then accuracy in identifying key species is improved, but universality across various process conditions deteriorates
Solution Approach 1:
The patent replaces expert-driven screening with an automated self-service system that uses low-dimensional simulations and graph-based importance analysis to automatically identify critical chemical reactions and key species. This self-service approach objectively evaluates reactions based on their actual influence in the specific process conditions being simulated, eliminating the need for expert pre-screening and enabling universal application across different processes and conditions without relying on subjective expert knowledge.
Data Source
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 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 the final chemical species DB.