Material Design Support Using Segmented Prediction Space
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The existing methods for designing metallic materials require significant computational resources and time due to the large number of design conditions, making efficient optimization challenging.
Innovation Solution
A material design support apparatus and method that includes setting design conditions, generating exhaustive prediction points, inputting them into a trained model to predict characteristic values, and adjusting design conditions based on the predictions to efficiently optimize material properties.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If optimization calculations are performed using a computer to reduce human and time burden, then design efficiency is improved, but computational load and calculation time increase
Solution Approach 1:
The patent segments the design condition space into multiple regions based on material characteristics and classification information. By dividing the exhaustive search space into manageable segments, the system can process each region separately, reducing the computational burden while maintaining comprehensive coverage of the design space.
Solution Approach 2:
The patent performs preliminary classification of design conditions by material characteristics before optimization calculations. This preliminary action organizes the design space in advance, allowing subsequent optimization to focus on specific material categories rather than performing exhaustive searches across all possible conditions, thereby reducing calculation time.
2Measurement precision
If the number of design conditions is increased to achieve comprehensive optimization, then design accuracy is improved, but computational resources and calculation time increase
Solution Approach 1:
The patent divides the large design condition space into multiple segments based on material classification information. Each segment can be processed independently with appropriate optimization calculations, achieving comprehensive coverage without requiring all computational resources to be allocated to a single exhaustive search.
Solution Approach 2:
The patent applies different optimization approaches and calculation methods to different material segments based on their specific characteristics. This allows the system to use computationally intensive methods only where necessary while using simpler methods for other segments, optimizing the balance between accuracy and resource consumption.
3Reliability
If exhaustive search is performed across all design conditions to ensure complete optimization, then solution completeness is improved, but calculation time increases significantly
Solution Approach 1:
The patent performs preliminary classification and filtering of design conditions based on material characteristics before conducting optimization searches. This preliminary action eliminates obviously suboptimal conditions early in the process, ensuring that the exhaustive search focuses only on promising candidates, thereby maintaining solution completeness while reducing calculation time.
Solution Approach 2:
The patent segments the exhaustive search into multiple parallel searches across different material classification segments. This allows the system to maintain complete coverage of the design space while distributing computational work across multiple smaller tasks that can be processed more efficiently and potentially in parallel.
Data Source
AI summary
A material design support apparatus for supporting optimization of a design condition for a material, the material design support apparatus including: a design condition setting unit configured to set a range of a design condition for a material; a required characteristic setting unit configured to set a range of a required characteristic of the material; an exhaustive prediction point generation unit configured to generate a plurality of exhaustive prediction points within the range of the design condition; a prediction unit configured to input the exhaustive prediction points into a trained model in which a correspondence between a design condition for the material and a characteristic value of the material is learned, thereby to predict a characteristic value of the material; and a design condition adjustment unit configured to adjust a range of the design condition for the material in which a plurality of exhaustive prediction points are subsequently generated.