Key Technologies For Big Data Analysis Of Dyeing And Heat Setting Quality

In the dyeing and finishing industry,dyeing and heat setting is the most important stage in product production,and it is also the stage that is most likely to cause high quality rework due to unqualified quality.Quality prediction and analysis is the key to achieving closed-loop quality control in the dyeing and finishing process.At present,quality prediction and analysis are not based on a large amount of data,thus,the accuracy of the results cannot be accurately determined regardless of the quality prediction or the analysis of the quality influence factors.In the actual heat setting,there are usually a plurality of unqualified quality indicators simultaneously.However,in the current analysis of influencing factors,the analysis of the single quality indicators is performed,and the improved solution cannot solve such problems.However,the traditional empirical methods in the dyeing process cannot accurately confirm the causes of chromatic aberration and cylinder difference,and there is no comprehensive consideration of the possible factors affecting the dye quality analysis.To solve the above problems,the research contents of this thesis are as follows:1.Based on small data and large amount of data,the relationship between quality index and influencing factors of heat setting process are further studied,a prediction model for the weight of process parameters on the quality index is established,and the accuracy of the prediction model in two cases are verified and compared.2.A large number of abnormal quality products are classified into different categories according to different quality indicators by cluster analysis.Combined with the correlation method,the influence of the process parameters on the quality of the corresponding different types was obtained.It can be realized on the big data platform,and can give reasonable improvement solutions.3.Based on the principle of the dyeing process,the actual dyeing production is studied.The factors that may cause chromatic aberration and cylinder differential conditions are comprehensively determined.The network structure of the dyeing process is constructed and simplified by the Bayesian network combined with the method of chain variable elimination.On the big data platform,the factors affecting the chromatic aberration and the cylinder difference are analyzed,and the two unqualified phenomena are predicted.4.According to the dyeing and finishing production process in the actual enterprise,build a dyeing and finishing big data platform.Design dyeing and finishing quality management system to make functions such as data collection,quality prediction and quality analysis more intelligent and visualized.Based on the above study,the accuracy and stability of the quality prediction and influencing factors analysis of the dyeing process and the heat setting process are improved.The product quality is improved,and the rework rate in production is reduced.It also provides effective reference value for realizing closed-loop control of dyeing and finishing quality and promoting the application of big data in the dyeing and finishing industry.