Relationship Between Fiber Properties and Rotor Spun Yarn Strength

摘要

The relationship between fiber properties-and yarn properties has been the focusof research, and considerable success has been achieved. Many mathematical modelshave been used to understand and predict the complex relationships between fiberparameters and yam characteristics, and substantial research has been done todetermine methods of predicting yarn properties. But the regression approach hasbeen used more intensively for prediction of the yarn strength with an assumption oflinearity. Recently, some studies have showed that relationship between yarn strengthand fiber properties is nonlinear. In this study we present a comparison study of three models for predicting thestrength of rotor spun cotton yarns from fiber properties. The adaptive neuro-fuzzyinference system (ANFIS) and Support Vector Machines (SVM)，generally calledArtificial Intelligent Techniques, that are capable of mapping non linear relations andMultiple Linear Regression models are used to predict the rotor spun yarn strength.HVI (high volume instrument) and Uster AFIS (advanced fiber information system)fiber test results are used to train and test the three models. Three important stageshave been involved in this study. In the fast stage, we used the ANFIS method to predict rotor spun yarn strengthform fiber properties; that is, we identify the relationship between yarn strength andfiber properties. The impact of each fiber property on the rotor spun yarn strength hasbeen analysed. The graphs illustrating the relationship between yam strength and oneof the fiber properties, with all the other properties held constant have been plotted.An examination of each graph revealed the nonlinear relationship between rotor spunyarn strength and all fiber properties .The direction of each fiber property influence onyarn strength has been also expected. Increasing fiber strength, upper half meanlength, length uniformity and yarn count has a positive impact whereas increasingmicronaire, yellowness and short fiber content has a negative impact on rotor spunyarn strength. impacts of fiber properties are non-linear. The study showed also how to control the yam quality using the knowledge on fiber properties through the yarnstrength leaned surfaces on fiber properties. In the second stage, we used Support Vector Nfachine (SVM) model to predictthe rotor spun yam strength. Using the SVM results, we have demonstrated thereiative importance of each fiber property on rotor spun yam strength. It wasconfirmed that all fiber properties ptay a role in performance of rotor spun yarnstrength. The results show that the rotor spun yarn strength is influenced to a greateror lesser degree by the fiber properties. Finally, the predictive performances of the two models are estimated andcompared to those from classical linear regression method. The comparison showedthat the results provided by ANFIS are better than those provided by both regressionand SVMs methods. The comparison between ANFiS and SVM with conventionalmethods indicated that these new approaches worked better in prediction of rotor yarnstrength and provided a good understanding of the nonlinear reiationship betweenfiber properties and rotor spun yarn strength.

目录

英文文摘

声明

Acknowledgement

Chapter 1. Introduction

Chapter 2. Literature Review

2.1 Rotor Spinning System

2.1.1 Introduction

2.1.2 Rotor Spinning Mechanism

2.1.3 Rotor Spun Yam Structure

2.1.4 Advantages of Rotor Spun Yarns

2.1.5 Disadvantages of Rotor Spun Yarns

2.1.6 End Use of Rotor Spun Yarns

2.1.7 Important Fiber Properties for Rotor Spinning

2.2. Previous studies on rotor spun yam strength and fiber properties

2.2.1 Theoretical and Experimental

2.2.2 Mathematical Models and Statistical Regression Methods

Chapter 3. Methodology

3.1 Introduction of Artificial Neural Network

3.1.1 Processing units

3.1.2 Connection between elements

3.1.3 Activation and output rules

3.1.4 Network topologies

3.1.5 Training of artificial neural networks

3.1.6 Leaning algorithms

3.2. Fundamental of Fuzzy logic

3.2.1 Fuzzy Logic

3.2.2 Fuzzy rules and Fuzzy Inference systems

3.2.3 Sugeno fuzzy model

3.3. Neural Fuzzy Systems

3.3.1 General Comparisons of Fuzzy Systems and Neural Networks

3.3.2 Adaptive Networks

3.3.3 Adaptive neuro-Fuzzy Systems Inference(ANFIS)

3.4. Support Vector machines

3.4.1 Introduction

3.4.1The Optimal Hyperplane

3.4.3 Feature Space

3.4.4 Kernel Functions

3.4.5 Support Vector Regression

3.4.6 Conclusion

Chapter 4. Rotor Spun yarn Strength Prediction

4.1 Introduction

4.2 Data Collection

4.3 ANTIS

4.3.1 Training Practical Considerations

4.3.2 Training Results

4.3.3 Analyzing of the Impact of fiber Properties on Rotor Yarn Strength

4.4 Support Vector machines(SVMs)

4.4.1 Practical considerations

4.4.2 Results Analysis

4.4.3 Importance of selection of the optimization methods and number of fold cross-validation

4.4.4 Importance of fiber properties on rotor spun yarn strength

4.5 Multiple linear Regressions

4.5.1 Practical Considerations

4.5.2 Results

4.6 Comparison Analysis

Chapter 5 Conclusions

5.1 Summary

5.2 Recommendation for future works

References

Appendix