Battery Laminate Abnormality Models for Precise X-Ray CT Inspection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing X-ray CT imaging methods for all-solid-state batteries have low resolution and limited element analysis, making it difficult to precisely estimate abnormalities within the battery laminate.
Innovation Solution
A model generation method using machine learning with X-ray image data and additional data from cut surfaces of battery laminate samples, including information from X-ray and ion beam observations, to create an abnormality estimation model.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If X-ray CT imaging is used to observe battery laminate, then non-destructive measurement is achieved, but measurement precision deteriorates due to low resolution
Solution Approach 1:
The patent combines X-ray CT imaging data with SEM image data and element analysis data into a unified teaching dataset. By merging multiple data sources with different strengths (X-ray CT provides non-destructive 3D structure, SEM provides high-resolution surface details, element analysis provides compositional information), the system creates a comprehensive model that achieves both non-destructive measurement and high precision abnormality estimation.
Solution Approach 2:
The patent introduces an abnormality estimation model trained with multi-source teaching data as an intermediary between direct observation and abnormality detection. This model acts as a mediator that translates low-resolution X-ray CT images into high-precision abnormality estimates by learning from correlated high-resolution SEM images and element analysis data, effectively bridging the resolution gap.
2Reliability
If X-ray CT imaging is used to observe battery laminate, then non-destructive measurement is achieved, but measurement capability deteriorates due to limited element analysis
Solution Approach 1:
The patent merges X-ray CT imaging data with element analysis data from multiple sources into a unified teaching dataset. By combining structural information from X-ray CT with compositional information from element analysis, the system creates a comprehensive model that simultaneously provides non-destructive measurement and detailed element analysis capability.
Solution Approach 2:
The patent creates a multi-functional abnormality estimation model that can perform both structural abnormality detection (from X-ray CT) and element analysis (from combined data sources) through a single unified system. This universal model handles multiple types of information (structural, compositional, morphological) and can estimate various types of abnormalities simultaneously, achieving multi-functionality that overcomes the limitations of single-modality imaging.
3Measurement precision
If SEM imaging is used to obtain high resolution images, then measurement precision improves, but destructive measurement occurs
Solution Approach 1:
The patent uses SEM images and element analysis data from destroyed samples as teaching data to train an abnormality estimation model. The model learns from these high-resolution 'copy' datasets without requiring the actual test samples to be destroyed. Once trained, the model can estimate abnormalities in intact battery laminates by processing their X-ray CT images, effectively using copies for training while preserving originals for evaluation.
Solution Approach 2:
The patent performs destructive SEM imaging and element analysis in advance on sample sets to create teaching datasets. This preliminary action on sacrificial samples allows the system to learn high-precision abnormality characteristics without needing to destroy the actual test samples later. The preliminary destruction of training samples enables non-destructive testing of evaluation samples.
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 precise, non-destructive estimation of abnormalities in battery laminates, improving battery performance and yield, thus enhancing energy efficiency.
Implementation Method 1
X-ray image data obtained by irradiating X rays onto the sample which is fixed by a jig of column shape
Implementation Method 2
information obtained by cutting the sample by irradiating an ion beam while fixing the sample to the jig
Data Source
AI summary
An abnormality estimation system (1) includes: an input data reception unit (2) that receives X-ray CT image data (7D) of a test object (7) which is a battery laminate as input data; a teaching data storage unit (6) that stores the X-ray CT image data of a sample which is a battery laminate and abnormality data of the same sample as teaching data; a model generation unit (3) that generates an abnormality estimation model for a battery laminate by machine learning using the teaching data stored in the teaching data storage unit (6); a model estimation unit (4) that estimates an abnormality in the test object (7) from the input data received by the input data reception unit (2) using the abnormality estimation model generated by the model generation unit (3); and a display unit (5) that displays estimation results from the model estimation unit (4).