Electrode Stack Corner Detection Using 3D Imaging and Neural Nets

Resolve Bottlenecks,
Find Innovative Solutions
Generate Solutions

Solution Overview

Problem

Current methods for determining the placement accuracy of electrode sheets in electrode composite stacks are not fully automated, precise, or robust, often relying on specialist evaluation of 3D image data with poor signal-to-noise ratios, and do not provide effective correction measures for placement errors.

Innovation Solution

A computer-implemented method using 3D imaging and convolutional neural networks to accurately determine the position of corner regions in electrode composite stacks by generating corrected edge profiles and extrapolating corner positions, ensuring precise and automated corner detection without compression or stretching of electrode sheets.

Engineering Contradictions & Design Principles

VSEngineering Contradiction Analysis

1Manufacturing precision

If electrode sheet placement accuracy is increased to ensure complete coverage, then manufacturing precision is improved, but production speed decreases

Engineering Contradiction:
Improveelectrode sheet placement accuracyVSAvoidproduction speed
Core Design Contradiction:
Manufacturing precisionVSProductivity

Solution Approach 1:

The patent replaces manual specialist evaluation of 3D image data with an automated neural network system that processes CT scan images to determine electrode placement accuracy. This automation resolves the contradiction by enabling rapid, precise measurement without requiring slow manual analysis, thus maintaining high manufacturing precision while increasing production speed.

Inventive Principle:
Principle #28Mechanics substitution (Replace mechanical system)

Solution Approach 2:

The patent changes the measurement parameters by using 3D CT scan imaging with specific slice thicknesses (0.5-2mm) and automated neural network analysis rather than manual 2D image evaluation. This parameter change enables faster processing while maintaining or improving measurement precision, resolving the speed-accuracy tradeoff.

Inventive Principle:
Principle #35Parameter changes

2Productivity

If 3D image data acquisition speed is increased, then productivity is improved, but signal-to-noise ratio deteriorates

Engineering Contradiction:
Improveimage data acquisition speedVSAvoidsignal-to-noise ratio
Core Design Contradiction:
ProductivityVSMeasurement precision

Solution Approach 1:

The patent applies preliminary action by using a trained neural network system that has been pre-trained on大量 CT scan images to recognize and filter noise patterns. This pre-training enables the system to rapidly process high-speed acquisition images while maintaining high signal-to-noise ratio through learned noise discrimination, resolving the contradiction between acquisition speed and measurement precision.

Inventive Principle:
Principle #10Preliminary action

Solution Approach 2:

The patent creates a digital 3D model (copy) of the electrode stack from CT scan data, allowing multiple analyses to be performed on the copied data without requiring additional physical scanning. This copying approach enables rapid repeated measurements of the same high-quality 3D data, maintaining signal-to-noise ratio while achieving high productivity through virtual rather than physical re-scanning.

Inventive Principle:
Principle #26Copying

3Extent of automation

If automated evaluation system is implemented, then extent of automation is improved, but device complexity increases

Engineering Contradiction:
Improveevaluation automationVSAvoidsystem complexity
Core Design Contradiction:
Extent of automationVSDevice complexity

Solution Approach 1:

The patent implements a universal neural network system that performs multiple functions: it identifies electrode positions, determines placement accuracy, detects rotation, and generates correction recommendations. This multi-functionality reduces overall system complexity by consolidating what would otherwise require multiple separate devices into a single automated evaluation system.

Inventive Principle:
Principle #6Universality (Multi-functionality)

Solution Approach 2:

The patent introduces a coordinate system as an intermediary that standardizes the representation of electrode positions and orientations. This intermediary framework simplifies the automation by providing a common reference system that the neural network can process uniformly, reducing the complexity of handling diverse measurement data types.

Inventive Principle:
Principle #24Intermediary (Mediator)

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 robust, precise, and fully automated determination of electrode sheet positions, improving manufacturing efficiency and accuracy by correcting placement errors and optimizing the manufacturing process.

Implementation Method 1

3D imaging of the corner region of electrode sheets of the electrode stack in a recording area using an imaging method, in particular a computed tomography imaging method

Methodology Applied
Scientific EffectX-ray: X-Ray

Implementation Method 2

a computed tomography imaging method, such that a data set is generated which includes 3D positional information of the electrode sheets

Methodology Applied
Scientific EffectComputed tomography: Tomography

Data Source

PatentEP4404149A1Method for determining a position of a corner region of an electrode composite stack
Publication Date: 2024.07.24 POWERCO SE
  • EP4404149A1 patent drawingFigure 1~2
  • EP4404149A1 patent drawingFigure 3~5
  • EP4404149A1 patent drawingFigure 6

AI summary

The invention relates to a method for determining the position of corners of polygonal electrode sheets in at least one corner region of an electrode stack (ESV), comprising the steps of: - 3D imaging of the corner region of the electrode stack (ESV) in a recording area using an imaging method, such that a data set is generated which includes 3D positional information of the electrode sheets (A, K) in the corner region of the electrode stack (ESV) relative to a support (5, SD, SB) or a marker arranged in the recording area, - Determining from the data set a first and a second edge profile (20x, 20y) of the edges (21x, 21y) framing the corner region of each electrode sheet (A, K), wherein a position of the corner (E1, E2, E3, E4) of the respective electrode sheet (A, K) is determined based on the edge profile (20x, 20y). The electrode sheets and their edges are determined using a neural network system.