Analysis And Optimization Of Process Parameters Of Drying Machine Based On Data-Driven

Drying is one of the important processes in the process of cigarette making.Its purpose is to remove excess moisture from the filaments,and to adjust the gas flow in the process of drying the filaments,such as temperature increase and humidity increase,material drying and material cooling,etc.Degrees of temperature and the wall temperature and other process parameters are used to adjust the moisture content of the outlet of the wire to improve and control the quality of the tobacco.The drying process of the drying machine includes the coupling of physics,chemistry and other multi-fields and multi-disciplines.The process parameters have many characteristics such as multi-dimension and complex relationship,Therefore,the fitting accuracy of the function model between the drying process parameters and the moisture content of the silk filaments established by the traditional method is low,resulting in the unreliability of the optimization setting of the process parameters of the dryer.For this reason,aimed at the problem that the automatic control method is used to control the fluctuation of the water content of the leaf,and the traditional parameter optimization method cannot consider the global process parameters of the dryer,resulting in the problem of unreliable optimization settings,this paper proposes a data-driven optimization method for the process parameters of the dryer.The specific research content and main conclusions are as follows:(1)Analyze the drying process and principle of the drying process of the drying wire dryer,and study the influence of parameters such as the wall temperature,humidity,and circulating hot air speed on the moisture content of the leaf silk during the drying process of the drying machine,and at the same time,Pearson correlation analysis and principal component analysis were carried out on the data characteristics of the parameters.The data characteristics of the drying process parameters were studied.The analysis results showed that there are many dimensions,autocorrelation,and multiple disturbances between the parameters of the drying process of the dryer.The relationship between the characteristics of the complex,for this purpose put forward a data-driven drying machine process parameters optimization method;(2)Based on the data characteristics of the above-mentioned drying process parameters,a lightweight data-driven prediction model was established based on neural network and principal component analysis to establish the mapping relationship between drying process parameters and target parameters,and to improve the operation of traditional neural networks.Efficiency and prediction accuracy.Based on this,the optimization model of drying process parameters was established based on genetic algorithm,and the optimal values of various process parameters and target parameters were calculated based on the above lightweight data-driven prediction model;(3)Through the actual production data of a company,it is verified that the data-driven drying process parameters optimization method proposed in this paper can achieve better target parameter prediction and optimization of process parameters under the conditions of more process parameters and complex relationships.The method improves the limitation of the setting and optimization of the process parameters of the dryer in the current production by personal experience,and provides a new way for accurately controlling the moisture content of the silk after drying and maintaining the stability of the quality of the product.