Research on the Digital Prediction Model of Cigarette Filament Feeding Process Based on BP Neural Network

In order to further achieve the moisture content at the outlet of the feeding process and stabilize the moisture content after cutting, and improve the production line process level, this paper studies a digital prediction model for the cigarette silk feeding process. Based on data and various algorithmic models, a set of process parameter control modeling methods is formed. Establish a digital prediction model for the outlet moisture content of the feeding process, and obtain the minimum error range trained by the Gradient Boosting Decision Tree (GBDT) algorithm. The trend of the predicted value is basically consistent with the actual value. The prediction model for the outlet moisture content of the feeding process can well fit the actual situation. Substitute the test set data for validation and trial operation, with an error range of -0.0906 similar to 0.0942. Based on the prediction model of the Gradient Boosting Decision Tree (GBDT) algorithm, the outlet moisture content of the feeding process has a small error and error range, and the distribution and trend of the predicted values and error values are uniform. The method used in this article for predicting the moisture content at the outlet of the feeding process has good accuracy, stable testing results, and guiding significance. Based on digital prediction, adjust parameters in a timely manner, enhance scientific judgment, reduce complex communication and adjust process data through experience, in order to improve process quality.