Welding state recognition method and device based on dual-channel features and electronic device

By acquiring the dual-channel force signal of the welding socket in real time and extracting and optimizing its frequency domain statistical features, the problem of not being able to fully inspect the welding status in lithium battery production in existing technologies has been solved, and the real-time monitoring and accurate identification of the welding status has been achieved.

CN116738141BActive Publication Date: 2026-06-09SBT ULTRASONIC TECH CO LTD
View PDF 2 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN Β· China
Patent Type
Patents(China)
Current Assignee / Owner
SBT ULTRASONIC TECH CO LTD
Filing Date
2022-12-29
Publication Date
2026-06-09

AI Technical Summary

Technical Problem

Existing offline quality monitoring cannot perform full inspection of the ultrasonic welding status in lithium battery production, resulting in inaccurate real-time monitoring of the welding status.

Method used

By acquiring the dual-channel force signals of the welding base in real time, including horizontal friction force signals and vertical pressure signals, frequency domain statistical feature extraction and optimization are performed to select the main features, which are then input into the welding condition monitoring model to achieve real-time acquisition of the welding condition.

Benefits of technology

It enables real-time full inspection of the welding status, ensuring accurate control of the welding status of each workpiece and avoiding the shortcomings of offline quality monitoring.

✦ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN116738141B_ABST
    Figure CN116738141B_ABST
Patent Text Reader

Abstract

Embodiments of the present application disclose a welding state recognition method and device based on double-channel features and an electronic device, and the present application relates to the technical field of ultrasonic welding. Instead of acquiring signals of workpieces, the present application acquires double-channel force signals of a welding seat in real time. At this time, frequency domain statistical features of the double-channel force signals are extracted to determine a plurality of frequency domain statistical features. The plurality of frequency domain statistical features are optimized to obtain relatively accurate main features. The main features are used for state classification and input into a welding state monitoring model to output a welding state in a welding process, so as to realize real-time acquisition of the welding state and facilitate full inspection of welding of the workpieces, thereby ensuring real-time grasping of the welding state of each workpiece and avoiding the problem that offline quality monitoring cannot perform full inspection on products.
Need to check novelty before this filing date? Find Prior Art