Filter Rod Pressure Drop Prediction Based on PSO-BP Neural Network

Filter rod pressure drop is a crucial parameter in filter rod production, which directly affects the quality of filter rod and the tasting experience, and the pressure drop is affected by many factors in the filter rod production process. In this study, we aim to solve the pressure drop prediction problem, and use BP neural network technology to predict the pressure drop of filter rods, optimize it by using particle swarm optimization (PSO), and construct a PSO-BP neural network model. In this study, an in-depth study on the relationship between pressure drop and multiple influencing factors was carried out, and a PSO-BP neural network-based model was established to explore the complex interaction between the pressure drop and different features in order to predict the trend of the pressure drop of the filter rod. The methodological steps include data collection and processing, feature selection, model training, and optimization of the model. The results of the study show that the proposed method exhibits significant accuracy and practicality in predicting the filter rod pressure drop. The results of the study show that the PSO-BP neural network is able to accurately predict the filter rod pressure drop in different cases, and it also performs well in terms of the mean relative error, the mean absolute error, the root-mean-square error, and the coefficient of determination of the samples, which provides a higher prediction accuracy.