Research on Moisture Prediction Model for Cigarette Drying Process

This article preprocesses real-time datasets, uses SPSS to analyze the correlation between various parameters in depth, and establishes mathematical relationship models between various parameters. Taking into account the influence of process parameters such as moisture content, temperature, and opening of the moisture outlet damper on the moisture content of the outlet, this paper chooses a neural network model based on LSTM to analyze and regulate the dynamic characteristics of the silk drying process, and evaluates the model through the R2 index. To improve the stability of moisture content in export drying silk, adjust the proportion and learning rate of training and testing samples, and select the optimal model based on the prediction accuracy of the modeling method. The results showed that the neural network model based on LSTM achieved correct prediction of drying process parameters, and the accuracy of controlling the moisture content at the drying outlet was significantly improved; This model has strong data analysis and production prediction capabilities, which not only meet the basic requirements of tobacco industry production, but also can be used for moisture prediction in various key links of cigarette production, and has certain practical application value.