Material Data Modeling Using Magnetization Heating-Cooling Features
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
The reliability of structure data obtained through measurement and observation in materials science is highly dependent on the skill of the person performing the measurement, affecting the accuracy of machine learning-based estimations in materials development.
Innovation Solution
A material data processing device and method that uses a computer to perform machine learning with process data, composition data, characteristics data, and microstructure data, including feature amounts based on magnetization temperature dependence during heating and cooling, to create a regression model and select appropriate feature amounts for estimation, thereby improving data acquisition and estimation precision.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If structure data are obtained through measurement and observation using conventional methods, then the data can be acquired, but the reliability is highly dependent on the skill of the person performing the measurement
Solution Approach 1:
The patent replaces manual measurement and observation methods with automated image processing and machine learning algorithms. Specifically, it uses automated extraction of microstructure data from images through computer-based analysis, eliminating the need for manual measurement while improving reliability and reducing dependence on personal skill.
Solution Approach 2:
The system enables self-service through automated data processing where the computer automatically extracts microstructure information from images using algorithms and machine learning models, without requiring manual intervention or expert observation skills for each measurement.
2Measurement precision
If manual measurement and observation methods are used for obtaining structure data, then data acquisition is possible, but the process is time-consuming and requires high skill
Solution Approach 1:
The patent replaces time-consuming manual measurement processes with automated computer-based image processing and machine learning algorithms that can rapidly analyze microstructure data without requiring expert intervention for each sample.
Solution Approach 2:
The system performs preliminary automated processing of microstructure images to extract relevant features before machine learning analysis, preparing the data in advance for efficient estimation and reducing overall processing time.
3Reliability
If feature amounts based on both heating and cooling magnetization temperature dependence are used, then more comprehensive data is obtained, but the complexity of data processing increases
Solution Approach 1:
The patent implements a dynamic approach where the machine learning model can adaptively select and process different types of microstructure data (heating, cooling, or both) based on the specific estimation task requirements, optimizing the balance between data comprehensiveness and processing complexity.
Solution Approach 2:
The system changes processing parameters by allowing flexible selection of which magnetization temperature dependence data (heating, cooling, or both) to use in the machine learning model, adjusting the level of data comprehensiveness according to the specific application needs.
Applied Scientific Principles
This section explains which scientific principles are used to turn an abstract innovation direction into a practical engineering solution.
Function Achieved in This Case
Enables easy data acquisition and high-precision estimation by reducing the reliance on personal skill and subjectivity in data collection, enhancing the accuracy of material property predictions.
Implementation Method 1
the microstructure data includes a feature amount based on a magnetization temperature dependence during heating, and a feature amount based on a magnetization temperature dependence during cooling
Data Source
AI summary
A material data processing device using a computer is provided with a regression model creation processing unit that performs machine learning using, out of process data, composition data, characteristics data, and microstructure data, two or more data including the structure data, and creates a regression model representing a correlation between respective data, an estimation processing unit that estimates, by using the regression model, the process data, the composition data, the characteristics data, or the microstructure data, having been used for machine learning, wherein the microstructure data includes a feature amount based on a magnetization temperature dependence during heating, and a feature amount based on a magnetization temperature dependence during cooling; and a temperature type selection means that selects use of either one or both of the feature amount based on the magnetization temperature dependence during heating or the feature amount based on the magnetization temperature dependence during cooling as the microstructure data to be used for the machine learning. The material data processing method includes performing the machine learning, and creating the regression model, selecting the use of either one or both of the feature amount based on the magnetization temperature dependence during heating or the feature amount based on the magnetization temperature dependence during cooling as the microstructure data to be used for the machine learning.