Prediction method of carding process production quality based on digital twin technology

Abstract

In recent years, digital twin technology has developed rapidly, which has great advantages in improving production quality and saving production costs. Therefore, it is of great significance to apply digital twin technology to the spinning field to realize the prediction of spinning production quality. This study takes the carding process in yarn manufacturing as an example. First, a prediction model framework for carding process production quality is proposed based on the digital twin modeling mechanism. Each key component’s modeling principles and methods are analyzed, and the modeling process and coupling between models are clarified. Next, the carding process production quality prediction model is established based on the proposed modeling framework and actual working conditions. Then, a set of comparative experiments on a generative adversarial network and a set of comparative experiments between different prediction models were designed. The results showed that the production quality prediction model of the carding process established according to the modeling method proposed in this paper meets the requirements of spinning production quality prediction, and the average error rate of the model was 5.7%. Finally, the modeling method and experimental results are discussed, and future research directions are considered. The paper has great significance for further research on quality prediction of yarn manufacturing, providing a reference for applying digital twin technology and promoting intelligent manufacturing in the spinning field.