Phase Map Generation via Principal Component Analysis
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing phase analysis methods require multiple steps and user expertise to select appropriate elemental combinations, making it difficult for inexperienced users to generate phase maps from elemental map data, especially when dealing with a large number of elements.
Innovation Solution
A phase analyzer and method utilizing principal component analysis to calculate scores, generate scatter diagrams, detect peak positions, classify points, and create phase maps, simplifying the process by automating the selection of elemental combinations and correlation determination.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Ease of operation
If traditional scatter diagram generation method is used, then phase map can be generated, but user must manually select element combinations from many elements which increases operation difficulty and time consumption
Solution Approach 1:
The system automatically performs principal component analysis and generates scatter diagrams using all available elemental map data without requiring manual user selection. The computer executes the analysis independently, allowing the system to serve itself by automatically determining optimal element combinations through mathematical transformation rather than relying on user expertise for selection.
Solution Approach 2:
The invention transforms the original elemental data parameters through principal component analysis, converting multiple elemental concentrations into principal component scores. This parameter transformation automatically identifies the most significant variations in the data, enabling the system to select optimal element combinations mathematically rather than through manual user judgment.
2Loss of information
If all elemental map data is analyzed, then comprehensiveness of phase distribution is improved, but data processing complexity and computational load increase
Solution Approach 1:
Principal component analysis transforms the original high-dimensional elemental data into a reduced set of principal component scores that capture the most significant variations. This mathematical transformation maintains the essential information about phase distributions while reducing computational complexity by focusing on the most informative dimensions of the data.
Solution Approach 2:
The system extracts the most significant information from the complete elemental map data by identifying and utilizing only the principal components that contribute most to the variance. This extraction process separates the essential phase distribution information from the redundant data, maintaining completeness while reducing processing burden.
3Measurement precision
If manual element combination selection is required, then analysis precision can be controlled, but time consumption and operational difficulty increase significantly
Solution Approach 1:
The computer system automatically performs the element selection and combination optimization through principal component analysis without requiring manual user intervention. The system independently determines the optimal element combinations by analyzing the variance and covariance structures of the data, eliminating the time-consuming manual trial-and-error process while maintaining analytical precision.
Solution Approach 2:
The system performs preliminary principal component analysis to pre-determine the optimal element combinations and their relationships before generating the final phase map. This preliminary computational action identifies the most informative variable combinations in advance, saving significant time during the actual phase analysis process.
Data Source
AI summary
A phase analyzer includes a principal component analysis section that performs principal component analysis on elemental map data that represents an intensity or concentration distribution corresponding to each element to calculate a principal component score corresponding to each unit area of the elemental map data, a scatter diagram generation section that plots the calculated principal component score to generate a scatter diagram of the principal component score, a peak position detection section that detects a peak position from the scatter diagram, a clustering section that calculates a distance between each point and each peak position within the scatter diagram, and classifies each point within the scatter diagram into a plurality of groups based on the distance, and a phase map generation section that generates a phase map based on classification results of the clustering section.