Battery Tab Defect Detection Before and After Blue Adhesive Pasting
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Current defect detection methods for lithium battery tabs during ultrasonic tab welding suffer from low accuracy, affecting the quality and safety of the cells due to issues like welding mark quantity defects, tab folding, and blue adhesive coverage defects.
Innovation Solution
A method that involves obtaining cell pictures before and after adhesive pasting, determining specific detection regions based on defect types, and performing defect detection on these regions to improve accuracy, including the use of deep learning models for efficient defect identification.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Productivity
If defect detection is performed after blue adhesive pasting, then the detection process can be completed in one step, but the detection accuracy decreases due to interference from adhesive color and reflection
Solution Approach 1:
The detection process is divided into two separate stages: first detecting tab defects before adhesive pasting, then detecting adhesive defects after pasting. This segmentation allows each detection to focus on specific defect types without interference from the other, resolving the contradiction between efficiency and accuracy by making the detection process modular and targeted.
Solution Approach 2:
Tab defect detection is performed preliminarily before adhesive pasting. By completing the tab detection earlier in the process, the system eliminates the need to distinguish tab defects from adhesive interference during a single combined detection, thereby improving accuracy while maintaining overall process efficiency through staged execution.
2Reliability
If conventional defect detection is performed after adhesive pasting, then all defects can be detected in one inspection, but the detection accuracy is low due to interference from adhesive color and reflection
Solution Approach 1:
The detection system segments defect types into tab defects (detected before adhesive pasting) and adhesive defects (detected after pasting). This segmentation ensures comprehensive coverage of all defect types while maintaining high accuracy for each category by eliminating the interference problem that would exist in a combined detection approach.
Solution Approach 2:
The harmful factor of adhesive color and reflection interference is extracted and removed from the tab defect detection process by performing that detection before adhesive application. This extraction eliminates the source of interference, allowing for accurate tab defect identification without the confounding factors present in post-adhesive detection.
3Reliability
If blue adhesive is pasted after ultrasonic welding, then the cell is protected and sealed, but interference from adhesive color and reflection reduces defect detection accuracy
Solution Approach 1:
Tab defect detection is performed as a preliminary action before adhesive pasting, when no interference exists. This timing allows for accurate detection of tab defects while the adhesive serves its protective function afterward, resolving the contradiction by sequencing the detection and protection functions in an optimal order.
Solution Approach 2:
The system maintains continuous useful action by performing tab defect detection before adhesive pasting (when detection accuracy is high) and then proceeding with adhesive application for cell protection. The useful actions of detection and protection are executed in sequence, each optimized for its specific function without compromising the other.
Data Source
Figure 1~3
Figure 4~6
Figure 7~9
AI summary
The present application relates to a defect detection method, a defect detection system, a defect detection apparatus, a device, and a storage medium. The method includes: obtaining a first cell picture of a to-be-detected cell before adhesive pasting and a second cell picture of the to-be-detected cell after adhesive pasting; determining a size of a cell detection region in the first cell picture according to a type of a to-be-detected defect and related parameters of the to-be-detected cell, and determining a position of the cell detection region in the first cell picture according to the size of the cell detection region and the related parameters of the to-be-detected cell; determining a size of a blue adhesive detection region in the second cell picture according to the type of the to-be-detected defect and the related parameters of the to-be-detected cell, and determining a position of the blue adhesive detection region in the second cell picture according to the size of the blue adhesive detection region and the related parameters of the to-be-detected cell; and performing defect detection on the electrical detection region and the blue adhesive detection region to obtain a detection result. The foregoing method can improve defect detection accuracy to some extent.