Device, method and program for acquiring feature data for material composition information based on artificial intelligence

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

The development of positive electrode materials for lithium-ion batteries is hindered by the lack of sufficient data for effective research using artificial intelligence, particularly deep learning, due to insufficient material data availability, making it difficult to analyze battery materials.

Innovation Solution

A device and method utilizing two AI models, one for composition information and one for structure information, with mutual learning and adjustment to enhance feature data extraction, enabling accurate feature data acquisition even with limited structural data.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Productivity

If artificial intelligence is used to analyze battery materials, then research speed and efficiency are improved, but the lack of sufficient material data limits the effectiveness of AI models

Engineering Contradiction:
Improveresearch speedVSAvoiddata sufficiency
Core Design Contradiction:
ProductivityVSLoss of information

Solution Approach 1:

The patent introduces a contrastive learning mechanism as an intermediary that bridges the gap between insufficient material data and AI model requirements. By using structure information as a mediator to guide the learning of composition information, the system can effectively train models even when direct material data is limited, thus resolving the contradiction between research speed improvement and data sufficiency limitations

Inventive Principle:
Principle #24Intermediary (Mediator)

2Ease of manufacture

If only composition information is used for AI analysis, then data acquisition is simplified, but the accuracy of material analysis deteriorates due to lack of structure information

Engineering Contradiction:
Improvedata acquisition easeVSAvoidmaterial analysis accuracy
Core Design Contradiction:
Ease of manufactureVSMeasurement precision

Solution Approach 1:

The patent segments the material information into two distinct components: composition information and structure information. By processing these segments separately and then integrating them through contrastive learning, the system maintains the ease of data acquisition from composition while incorporating structure information to enhance analysis accuracy, thus resolving the contradiction between data acquisition simplicity and analysis precision

Inventive Principle:
Principle #1Segmentation

3Device complexity

If deep learning models are trained with limited data, then model complexity is reduced, but the predictive performance and reliability deteriorate

Engineering Contradiction:
Improvemodel complexityVSAvoidpredictive performance
Core Design Contradiction:
Device complexityVSReliability

Solution Approach 1:

The patent merges two different types of information (composition and structure) into a unified learning framework through contrastive learning. This combination allows the model to leverage complementary information from both sources, improving predictive performance and reliability without requiring excessive model complexity or large amounts of data, thus resolving the contradiction between model simplicity and predictive reliability

Inventive Principle:
Principle #5Merging (Combining)

Data Source

PatentEP4546354B1Device, method and program for acquiring feature data for material composition information based on artificial intelligence
Publication Date: 2026.02.25 LG MANAGEMENT DEV INST CO LTD
  • EP4546354B1 patent drawingFigure 1~2
  • EP4546354B1 patent drawingFigure 3~4
  • EP4546354B1 patent drawingFigure 5

AI summary

A system, device, method, and program for acquiring feature data for material composition information based on artificial intelligence are disclosed. The system may include a memory configured to store a first artificial intelligence (AI) model configured to output first feature data for composition information of a material and a second AI model configured to output second feature data for structure information of the material; and a processor configured to learn the first AI model and the second AI model. The processor may be configured to learn the first AI model based on the second feature data for the structure information of the material output by the second AI model, and/or to learn the second AI model based on the first feature data for the composition information of the material output by the first AI model.