Assessing the manufacturing of a textile body using color parameters
Find Innovative SolutionsGenerate Solutions
Solution Overview
Problem
Existing methods for assessing yarn quality, particularly in elongate textile bodies, are inefficient and lead to time and material loss due to the removal of non-defective sections during the rewinding process.
Innovation Solution
A method and apparatus that utilize principal component analysis to assess yarn quality by determining deviation scores based on color and non-optical parameters, weighing deviations differently depending on their statistical dispersion, and using a control unit to improve the robustness of quality assessment.
Engineering Contradictions & Design Principles
Engineering Contradiction Analysis
1Measurement precision
If yarn quality assessment uses traditional parameter monitoring methods, then the assessment process is simple, but the accuracy is insufficient leading to removal of non-defective sections
Solution Approach 1:
The patent transforms the quality assessment from monitoring individual parameters to analyzing principal components derived from multiple parameters. This dimensional transformation enables more accurate identification of defective sections by capturing correlations between parameters, thereby improving measurement precision without requiring overly complex additional hardware
Solution Approach 2:
The patent introduces principal components as intermediary variables that mediate between raw sensor measurements and quality assessment decisions. These principal components serve as a bridge that consolidates information from multiple parameters, improving assessment accuracy while maintaining manageable system complexity
2Reliability
If all sections with parameter deviations are removed during rewinding, then yarn quality is improved, but time and material are lost
Solution Approach 1:
The patent applies local quality assessment by evaluating each section's deviation in the context of learned normal variations. Instead of applying uniform removal criteria, the system locally determines which deviations represent actual defects versus normal variations, preventing unnecessary removal of non-defective sections and reducing time loss
Solution Approach 2:
The patent implements feedback mechanisms where the system continuously learns from measured data to refine its understanding of normal parameter variations. This feedback loop enables more accurate distinction between defective and non-defective sections, improving yarn quality while minimizing unnecessary removals and associated time losses
3Reliability
If all sections with parameter deviations are removed during rewinding, then yarn quality is improved, but material is lost
Solution Approach 1:
The patent enables localized quality judgment by assessing deviations in the context of principal components that capture the essence of normal variations. This allows precise identification of actual defects versus acceptable variations, ensuring that only truly defective sections are removed, thereby maintaining yarn quality while minimizing material loss
Solution Approach 2:
The system uses feedback from continuous measurement and learning to refine its defect identification criteria. By learning from historical data about normal parameter variations, the system becomes more accurate in distinguishing defects from acceptable variations, reducing unnecessary material removal while maintaining high yarn quality standards
Data Source
Figure 1
Figure 2~4
Figure 5(A)~5(C)
AI summary
In order to assess the quality of an elongate textile body, such as a yarn precursor or a yarn, the textile body (52) is run past a sensor head (12, 18, 24, 30; 44). At a first plurality of times, body parameters of the textile body (52) are measured. The parameters can include color components and/or other parameters, such as the thickness of the textile body (52). Principal component analysis is used to scale deviations in parameter space to make the quality assessment more robust.