Research On Optimization Of Process Parameters Of Filter Rod Production Based On Data Mining

China's cigarette industry has a history of more than 100 years.At the beginning,70% of the cigarette market products were imported from abroad or produced by foreign capital in China.Later,China's "out of the zone" strategy was implemented,which shows China's tobacco industry has developed rapidly.Tobacco industry has become an important part of China's national economic system.The vigorous development of the tobacco industry has also made competition among cigarette companies increasingly fierce,and improving consumer satisfaction has become one of the key research tasks of tobacco companies.The quality of cigarettes is a key metric for measuring consumer satisfaction,so it is becoming increasingly important for companies to improve product quality.In a sense,product quality determines the core competitiveness of an enterprise.With the continuous advancement of the informationization process of tobacco companies,the quality management and production data accumulated in tobacco companies has rapidly increased.Enterprises hope to carry out deep-level mining of these data to provide assistance for enterprise management and decision-making.Faced with data of massive and low information density,traditional statistical analysis methods have not been able to solve the problem well.Businesses urgently need a new way to accomplish this task.Data mining is a technique for extracting hidden and potentially valuable information from massive amounts of data.The use of data mining technology to achieve the diagnosis and optimization of quality problems provides an effective means to improve the quality of enterprises' products and increase their competitiveness.Tobacco companies,as traditional production companies,have less research on data management in terms of quality management and optimization.With continuous improvement of people's health consciousness,the quality ofcigarette filter rods has been paid more and more attention in tobacco companies.Cigarette filter rods as one of the important raw materials for producing cigarettes,the stability of its quality plays a vital role in the stability of the overall quality of the cigarette.The suction resistance of the filter rod directly affects the overall taste of the cigarette,and also affects the adsorption effect of harmful substances.How to adjust the production process parameters in the production process to stabilize the absorption resistance of the filter rod near the set value and reduce production fluctuations has become a major problem that tobacco companies need to solve.In this paper,data mining technology is used to analyze the factors affecting the absorption drag fluctuations of filter rods,and the settings of filter rod production process parameters are optimized.Firstly,this article uses the real-time production data on the filter rod production line to propose two data processing methods: moving standard deviation and direct calculation of standard deviation.Calculate the standard deviation of the filter rod absorption resistance to describe the filter rod absorption resistance fluctuations.Then use machine learning methods to build a mathematical model that affects the absorption drag fluctuations of the filter rod for the two sets of data.Through model comparison,the data processed by moving standard deviation and the random forest method is finally selected for model construction.Differential analysis was made on the filter rod suction resistance fluctuations in four different situations:brand,tow specification,suction resistance setting standard and different production periods.Based on the genetic algorithm,two optimization schemes are proposed to predict the adjusted production process parameters,respectively.After verification,it was found that the random forest and genetic algorithm can optimize the filter rod production process parameters,reduce the filter rod absorption resistance fluctuation,improve the product quality,reduce the quality cost of the enterprise in the production process,and thereby improve the core competitiveness.