Electrode Winding Vision Control for Meandering Detection
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
In the manufacturing of cylindrical secondary batteries, meandering issues during the winding process lead to positional inaccuracies of electrodes and separators, resulting in safety risks and defective products due to subjective human judgment and intermittent inspection methods.
Innovation Solution
A manufacturing apparatus and method utilizing machine vision systems to measure non-overlap widths between electrodes and separators, with a controller processing vision data to compute and judge meandering defects, and a data processing unit accumulating and processing defect judgment data through machine learning for real-time automation.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Reliability
If meandering inspection is performed through sampling inspection or only when there is a change such as material replacement, then inspection cost and time are reduced, but meandering defects cannot be detected continuously and reliably
Solution Approach 1:
The patent implements continuous meandering inspection throughout the electrode assembly manufacturing process using machine vision systems. Multiple cameras capture images at different positions, and the control unit continuously processes these images to detect meandering defects in real-time, replacing intermittent sampling inspection with continuous monitoring.
Solution Approach 2:
The patent replaces manual inspection methods with an automated machine vision system. Cameras, control units, and image processing algorithms automatically detect and judge meandering defects, eliminating the need for workers to disassemble and visually inspect jelly-rolls, thereby improving both reliability and efficiency.
2Measurement precision
If meandering inspection is performed by disassembling jelly-roll and checking real thing by worker, then defect detection accuracy is improved, but productivity decreases and subjective judgment errors occur
Solution Approach 1:
The patent replaces manual disassembly and visual inspection with an automated machine vision system. Multiple cameras capture images of the electrode assembly during manufacturing, and the control unit processes these images to objectively detect meandering defects, eliminating subjective judgment errors and improving inspection speed.
Solution Approach 2:
The patent creates optical copies (images) of the electrode assembly using machine vision systems. Instead of physically disassembling the jelly-roll to inspect it, the system captures images at different positions and processes these copies to detect meandering defects, maintaining detection accuracy while improving productivity.
3Difficulty of detecting and measuring
If non-overlap width measurement is performed by X-ray or manual inspection, then meandering detection capability is improved, but measurement precision decreases due to subjective judgment
Solution Approach 1:
The patent replaces X-ray measurement and manual inspection with a machine vision system that uses cameras to capture images and automatically measures non-overlap widths. The control unit processes these images to objectively determine the non-overlap widths between electrodes and separators, eliminating subjective judgment and improving measurement precision.
Data Source
AI summary
A manufacturing apparatus of an electrode assembly includes first and second machine visions provided on driving-direction front ends of first and second electrodes from a winding core to measure first non-overlap widths between first width-direction end portions of the first and second electrodes and first width-direction end portions of first and second separators, respectively; a controller collecting vision data obtained from the first and second machine visions; setting a non-overlap width between the first separator and the second separator to 0 through vision data, computing a third non-overlap width between the first width-direction end portions of the first and second electrodes, and judging a meandering defect with the computed third non-overlap width; a data processing unit configured to accumulate and process defect judgment data through machine learning; and an input unit configured to automatically input a meandering reference value based on the data processed by the data processing unit.