A yarn tension variation detection method based on visual measurement

By using multi-exposure visual measurement and an improved Fourier neural operator model, the problems of insufficient accuracy in yarn tension detection due to friction and visual detection at high speeds were solved, achieving high-precision and stable detection of yarn tension changes.

CN122409031APending Publication Date: 2026-07-17JUANCHENG YONGCHUANG TEXTILE CO LTD
View PDF 0 Cites 0 Cited by

Patent Information

Authority / Receiving Office
CN ยท China
Patent Type
Applications(China)
Current Assignee / Owner
JUANCHENG YONGCHUANG TEXTILE CO LTD
Filing Date
2026-05-09
Publication Date
2026-07-17

AI Technical Summary

Technical Problem

Existing yarn tension detection methods are prone to friction and wear under high-speed operation and fine yarn conditions, and visual inspection has difficulty accurately capturing the span deformation characteristics at the moment of tension change, resulting in insufficient detection accuracy and stability.

Method used

By employing multi-exposure visual measurement and an improved Fourier neural operator model, the visual measurement span is determined between the yarn traction and winding positions. Multiple short-pulse illuminations are applied to construct the yarn span deformation field and perform phase response analysis. Combined with the span phase anchoring spectrum correction mechanism, the detection accuracy and stability are improved.

Benefits of technology

It effectively captures minute changes in sag and curvature caused by yarn tension variations, reduces the influence of motion ambiguity, improves the accuracy and anti-interference ability of yarn tension change detection, and enhances the applicability of continuous operation of the detection.

โœฆ Generated by Eureka AI based on patent content.

Smart Images

  • Figure CN122409031A_ABST
    Figure CN122409031A_ABST
Patent Text Reader

Abstract

The application discloses a yarn tension change detection method based on visual measurement, which comprises the following steps: step one: determining a visual measurement span and applying multiple short pulse illuminations to obtain a yarn multi-exposure visual frame group; step two: separating multiple yarn trajectories to obtain a yarn center line set at a sub-moment; step three: establishing a span physical coordinate mapping relationship to obtain a yarn span deformation field; step four: analyzing a phase response to obtain a yarn tension phase response diagram; step five: extracting a tension change fingerprint and mapping it into a tension state node; step six: constructing a tension visual field; step seven: identifying a tension change event by using an improved Fourier neural operator model; and step eight: screening, remapping and verifying the identification result to generate a yarn tension change detection result. The span phase anchoring spectrum correction mechanism and the improved Fourier neural operator model are used to realize high-precision visual detection of yarn tension change.
Need to check novelty before this filing date? Find Prior Art