中国烟草学报2025，Vol.31Issue(3)：60-69,10.DOI:10.16472/j.chinatobacco.2024.T0080
基于PSO-ELM的不同温湿度条件下叶丝干燥入口水分控制研究
Research on moisture control at the inlet of leaf fil drying under different temperature and humidity conditions based on PSO-ELM
摘要
Abstract
[Background and Objective]This study aims to control the quality of moisture at the inlet of tobacco fil drying under different temperature and humidity conditions,promoting the stability of the tobacco fil drying process and improving the quality of finished tobacco fil.[Methods]K-means clustering analysis was used to divide the temperature and humidity intervals.Statistical analysis was employed to distinguish the quality of moisture at the inlet of leaf fil drying under different temperature and humidity intervals.Classification models for moisture at the inlet of leaf fil drying under different temperature and humidity conditions were constructed,and optimal process parameters were selected based on the classification models.[Results](1)The year can be divided into four intervals:April and May as medium temperature and low humidity;June,July,and August as high temperature and high humidity;September and October as medium temperature and medium humidity;and other months as low temperature and medium humidity,with significant differences in moisture at the inlet of leaf fil drying under different temperature and humidity intervals;(2)After discretization,,the moisture at the inlet of leaf fil drying under different temperature and humidity intervals was classified into four quality categories:poor quality(others),medium quality(low moisture content μ-1.5 σ～μ-0.5 σ),high quality(μ-0.5 σ～μ+0.5 σ),and medium quality(high moisture content μ+0.5 σ～μ+1.5 σ);(3)T The PSO-ELM classification model for moisture at the inlet of drying under different temperature and humidity intervals outperformed GS-SVM and GS-RF,with accuracy,precision,and recall rates over 90％for each temperature and humidity interval,and F1 scores above 0.90;(4)After applying the PSO-ELM model to select process parameters that maximize high-quality moisture at the inlet of drying in actual production,the standard deviation of moisture at the inlet of leaf fil drying under different temperature and humidity conditions was reduced by 40％～50％,and the proportion of high-quality moisture significantly increased,with the proportions in the medium temperature and low humidity and low temperature and medium humidity intervals increasing by 38.9％and 60％,respectively.关键词
叶丝干燥/温湿度/粒子群/极限学习机Key words
leaf fil drying/temperature and humidity/particle swarm optimization/extreme learning machine引用本文复制引用
李自娟,陈娇娇,李宜馨,吕萱,赵海洋,孙朔,冯子贤,高杨,赵力源,呼守宇..基于PSO-ELM的不同温湿度条件下叶丝干燥入口水分控制研究[J].中国烟草学报,2025,31(3):60-69,10.基金项目
河北中烟工业有限责任公司科技项目"基于大数据分析的制丝过程稳态控制与应用技术研究"(HBZY2023A040)(AW201911)(2023130700341017) （HBZY2023A040）
