The Visual System For Yarn Of Ramie Quality Prediction Based On Machine Learning In Python

In the process of textile processing,there is a certain relationship between the fiber index of the raw material and the yarn quality index,we can use the neural network to predict the yarn quality.However,at present,China's woolen textile industry enterprises,the level of intelligence is not enough.For example,they still use the traditional manual methods to do the data analysis.Although there are many studies have shown that neural network has a high accuracy in the prediction of yarn quality.Ordinary textile enterprises can not be applied to the actual production processes using the neural network.To be honest,there is not a specific quality prediction system design for general ordinary workers in the general textile industry.In this situation,these predictive models fail to be customized.(1)Based on the previous research on the yarn quality prediction algorithm.Using the Ramie fiber index: the average breaking tenacity of flat bundles(CN/Tex),the average number of dry Ramie(Nm),degummed hard,combed granulite,combed sliver and yarn,combed sliver weight unevenness,sliver weight deviation,combing the moisture regain is trained to predict the yarn unevenness,yarn evenness,into three performance indexes of yarn granulite.Based on the MEA-BP prediction model,the LM algorithm is used to optimize the MEA-BP to improve the prediction accuracy and accelerate the convergence rate.The neural network algorithm based on MEA-LM-BP is proposed to predict the yarn forming model.The wavelet neural network algorithm is used to predict the quality of yarn.The results show that the wavelet neural network has high prediction accuracy in predicting the quality of yarn.(2)This paper designs the yarn quality prediction system based on the modular design of the Python platform.Including the prediction module,the data processing module and the visualization module,and provides a convenient GUI interface.Users only need to click the button,in this way,we can use the model to complete the operation of yarn quality prediction.(3)Also designed a simple online linen data management page to facilitate the online management of data to facilitate the relevant personnel can increase data,delete,change andother operations everytime and everywhere.The most important is that this web page can be directly with the yarn prediction system for docking and even real-time data can be achieved,this design is more in line with the trend of large data age.