AI Mixture Property Prediction with Molecular Features and Interactions
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional methods struggle to efficiently predict the physical properties of complex mixtures with multiple materials due to increasing complexity in mixing ratios, limiting their effectiveness.
Innovation Solution
An artificial intelligence-based device and method using AI models trained to extract feature data and predict physical properties, including attention-based and molecular contrastive learning models, to analyze mixtures with multiple materials.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If conventional methods are used to predict physical properties of mixtures, then the method is simple to implement, but the prediction accuracy deteriorates as the number of materials and mixing ratio complexity increases
Solution Approach 1:
The prediction system is segmented into multiple stages: (1) extracting molecular features from individual materials, (2) calculating interaction energies between material pairs, (3) aggregating interactions to predict mixture properties. This segmentation allows the system to handle complex mixtures by breaking down the prediction task into manageable components, resolving the contradiction between accuracy and complexity.
Solution Approach 2:
The patent introduces interaction energy as an intermediary parameter that mediates between individual material properties and mixture properties. By calculating interaction energies between material pairs and aggregating them, the system achieves accurate predictions for complex mixtures without requiring direct analysis of all possible material combinations, thus managing complexity effectively.
2Adaptability or versatility
If the number of materials in the mixture increases, then the application versatility improves, but the prediction efficiency deteriorates
Solution Approach 1:
The patent merges the prediction of multiple materials into a unified framework based on pairwise interactions. Instead of treating each material separately or requiring complete analysis of all combinations, the system combines individual material features with pairwise interaction energies to predict mixture properties. This merging approach maintains versatility for handling multiple materials while improving efficiency through systematic aggregation.
Solution Approach 2:
The system changes the parameter representation from detailed molecular-level descriptions of all materials to aggregated interaction energy parameters. By transforming the problem into one of calculating and aggregating interaction energies rather than analyzing complete molecular configurations, the system achieves both high versatility for complex mixtures and improved prediction efficiency.
Data Source
Figure 1~2
Figure 3~5
Figure 6a
AI summary
A device for predicting the physical properties of a mixture including a plurality of materials is disclosed. The device may comprise a memory in which a first AI model trained to output first feature data of material information, and a second AI model trained to output physical property prediction information of the first feature data are stored, and a processor for executing the first AI model and the second AI model, wherein the processor may input, into the first AI model, material information of each of the plurality of materials to acquire first feature data of each of the plurality of materials, and input the first feature data into the second AI model to acquire physical property prediction information of the mixture.