Study On Quality Prediction And Appearance Virtual Simulation Of Yarn

The quality of yarn directly affects the economic benefits of textile enterprises, so it is necessary to establish a sound, practical yarn quality prediction model and yarn appearance simulation model which can help designers design and modify the yarn product based on the quality indicators of raw materials and the process parameters, and provide an effective method for textile enterprises to reduce costs and ensure the quality of yarn.This paper establishes an yarn quality prediction model based on ant colony clustering algorithm with the research of the influence factors of yarn quality of the performance indicators, analyzes the reasons and shortages of cluster analysis of the basic ant colony algorithm in building yarn quality prediction model, and designs an improved yarn quality prediction model based on adaptive ant colony clustering algorithm. The results show that the model largely avoids the phenomenon of falling into local optimal solution, increases the convergence speed of prediction models and improve the system stability. The model has high prediction accuracy and can solve the problem of yarn quality prediction. In this paper, we have researched deeply on the appearance of morphological structure of the yarn and adopted two methods to establish the simulation model of yarn appearance: one is to establish simulation model using skeleton model based on the technique of the virtual reality of the model; the other is to adopt non-Uniform rational B-spline surface algorithm to establish simulation models. The simulation results can be seen that the establishment of simulation model of the yarn using the two methods all achieved relatively good results and reached the expectant targets, at the same time it also demonstrate the feasibility of the program.