Graph Neural Network for Droplet Behavior Prediction
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Conventional film forming techniques require extensive time and cost for adjusting the arrangement pattern of curable composition droplets and conditions for mold pressing, due to the need for repetitive physical calculations, especially when dealing with numerous combinations of droplet patterns, leading to inefficient simulation processes.
Innovation Solution
A prediction method using a learning model to predict the behavior of droplets, where the input includes information on droplet positions, employing a graph neural network to simulate the behavior and interaction of droplets, reducing the need for physical calculations and improving simulation efficiency.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If physical calculation is used to predict droplet behavior, then prediction accuracy is improved, but calculation time increases enormously
Solution Approach 1:
The system performs preliminary actions by pre-calculating droplet behavior through physical calculations during an offline training phase. These pre-calculated results are stored as a database that can be quickly queried during actual film formation processes, eliminating the need for real-time physical calculations while maintaining prediction accuracy.
Solution Approach 2:
The system creates a simplified copy of the complex physical calculation model by training an AI model on pre-calculated physical simulation data. This AI model copy can predict droplet behavior with sufficient accuracy but executes much faster than the original physical calculation model, resolving the contradiction between accuracy and speed.
2Manufacturing precision
If repetitive simulation is performed to adjust droplet arrangement patterns, then film formation quality is improved, but productivity decreases
Solution Approach 1:
The system implements feedback by using the AI model to predict droplet behavior for different arrangement patterns and providing this prediction information back to the control unit. This allows rapid evaluation of multiple patterns without repetitive physical simulations, improving both quality optimization and productivity.
Solution Approach 2:
The system changes the parameter representation by transforming complex physical simulation parameters into AI model input parameters based on droplet arrangement geometry. This parameter transformation enables rapid prediction across different arrangement patterns, improving adjustment efficiency while maintaining film formation quality.
Data Source
AI summary
A prediction method of predicting a behavior of droplets of a curable composition in a process of forming a film of the curable composition from a plurality of droplets of the curable composition arranged on a first member, the method including predicting the behavior of the droplets using a learning model, wherein an input of the learning model includes first information indicating positions on the first member to which the droplets of the curable composition are to be arranged.