A Diagnostic Expert System for the Coloration of Textile Fiber Blends
General Material Designation
[Thesis]
First Statement of Responsibility
Zubair, Muhammad
Subsequent Statement of Responsibility
Shamey, Renzo
.PUBLICATION, DISTRIBUTION, ETC
Name of Publisher, Distributor, etc.
North Carolina State University
Date of Publication, Distribution, etc.
2020
PHYSICAL DESCRIPTION
Specific Material Designation and Extent of Item
369
DISSERTATION (THESIS) NOTE
Dissertation or thesis details and type of degree
Ph.D.
Body granting the degree
North Carolina State University
Text preceding or following the note
2020
SUMMARY OR ABSTRACT
Text of Note
Materials based on fiber blends are among the most common textile products and they are projected to continue to expand. The coloration of textile fiber blends, however, is challenging when compared to single fiber materials. This is due to differences in the dyeing behavior of the fiber components of the blend. The colorant class and application conditions required for one fiber may damage or not be suitable for the other component in the blend. Th colorants may cross-stain the fiber they are not designed for, and moreover there is a possibility of physical and chemical interaction between different colorant classes and chemicals used during the process. These issues increase the complexity of the coloration of fiber blends. With an increasing demand for the rightfirst- time dyeing approach to reduce costs and remain competitive, the need to producing dyed materials that are free of faults and possess suitable levels of quality, at an affordable cost, is more apparent. Troubleshooting problems in the coloration of blends is a difficult task considering the large numbers of factors involved. The determination of the exact cause of the problem can be very difficult. Dyehouses usually utilize human experts to troubleshoot problems. Such experts are often able to considerably reduce the number of probable causes involved. In some cases, however, the determination of the actual cause of the problem requires a detailed analysis. Over the years the availability of human expertise in the domain of textile coloration has declined considerably and experts have become rather scarce. An Expert System is a computer program that can be used to mimic the knowledge and experience of human experts in troubleshooting problems in various domains. The goal of the present study is to develop a functional knowledge-based expert system for troubleshooting problems in the coloration of fiber blends (DEXPERT-B). The system was developed in three phases; Phase I involved the identification of common blends and knowledge acquisition. Phase II covered the design and development of the system and in Phase III testing and evaluation of the system were performed. Polyester/cellulosic (PES/CELL) blends are among the most common and widely used fiber blends in textiles. This blend type was identified and selected based on trade statistics and published literature. The most common problems in the coloration of these blends in different material forms (e.g. yarn, knitted and woven fabrics) with various colorants classes (pigment, disperse, reactive, direct, vat and sulfur dyes), utilizing different coloration process (batch, semicontinuous and continuous), methods (one bath, two bath) and corresponding coloration machines was reviewed in detail. This was achieved through a survey of technical literature and interviews with practical dyers. Based on the knowledge base gathered, an electronic survey comprising common coloration faults and their potential causes in the PES/CELL blends was developed. The survey was employed to acquire specialized knowledge from coloration experts. The experts' responses were analyzed and used to complement the knowledge base. The expert system shell was developed using wxCLIPS which is the modified version of the C Language Integrated Production System (CLIPS) to provide a custom GUI functionality. The system was verified, validated and evaluated using human experts. Several faulty dyed samples were obtained from production mill and used for the assessment of the comparative performance of human experts and expert system in identification of the type and root causes of faults. The test results showed good performance of the expert system when compared to human experts. This system is developed for use in actual production settings for the diagnosis of common coloration problems in the PES/CELL blends. With the aid of the expert system the number of probable causes for a particular problem can be isolated and reduced considerably. The developed system has a potential utility for use as an educational and training tool for novice and practical dyers. DEXPERT-B can help practical dyers in a quick resolution of problems by identifying causes and recommending solutions to problems which can help improve the efficiency of the coloration of PES/CELL blends.