Application of Statistical Control Charts for Monitoring the Textile Yarn Quality

Abstract

At the current time, textile product quality is the most attractive factor for the consumer market. Iraqis’ textile yarn industries are facing a lot of difficulties and competition of cotton yarn products, which has been increased versus artificial fibers. The main problems include the physical and chemical characteristics of cotton yarn because of genetic, environmental, harvesting, and ginning factors. The Statistical control process is a powerful and useful methodology used to solve problems in textile yarn industries to achieve process stability, improve process capability, and reduce process variability. The main objective of this study is to apply control charts and comparison control chart performance of yarn spinning data for quick detection of process shifts that occur to take corrective action. Therefore, in this study, the control charts have been applied at the Wasit state company for textile industries to control the quality of cotton yarns produced. Applying control charts in all yarn spinning stages is a very important issue, especially in production cost and yarn quality. Quality control charts selected for variables that include ( X ¯ -R and X ¯ -S charts) are constructed to describe tenacity, elongation, and the coefficient of variation. Thirty samples size of yarns with five reading of observations per sample of count 1/27Ne are drawn from spinning machine. Minitab is statistical software used to construct the control charts because of its good reputation which is confirmed by the results achieved in our research. The final results of this study will help us to distinguish yarns parameters in the points of the economy and the quality, by comparison, their above-mention parameters.

Prediction Model of Rotor Yarn Quality Based on CNN-LSTM

“…As shown in Figure 14, when predicting yarn strength, one layer of LSTM, two layers of LSTM, and three layers of LSTM were used, and the MSE value decreased only slightly. From the point of view of reducing computation, computing time, and saving resources, this model chooses one-layer LSTM [20][21][22] Momentum, and Adam optimization algorithms. These four optimal algorithms have their own characteristics, SGD (Stochastic Gradient Descent) trains for large samples; Adagrad (Adaptive Gradient Algorithm) is an improvement of SGD to improve its robustness; Momentum is also an improvement on SGD, which can accelerate the convergence of SGD and has a strong inhibition on convergence oscillation; Adam (Adaptive Moment Estimation) only needs to give an initial learning rate, and can adapt the learning rate according to the situation in the training process [23,24].…”

Kali̇te Kontrol Çi̇zelgeleri̇ni̇n Bi̇bli̇yometri̇k Anali̇zi̇

“…Suryoputro et al (2017) implemented seven tools including control charts at the Batik Company to reduce and prevent defects in textile products. Abdulghafour et al (2021) conducted a study to help textile manufacturers and enable them to focus on reducing quality costs in their relevant unit operations, especially in a competitive market or recession environment. 𝑋̅ -R and 𝑋̅ -S control charts for variables have been applied and implemented to monitor the yarn quality variations.…”