Construction and Application of Precise Prediction Model for Cigarette Weight of Cigarette Equipment

TANG Hanxi; ZHANG Sheng; XU Ran; LYU Xiaobo; CHEN Yanmeng; WU Yongjin; CHEN Wentao

doi:10.19554/j.cnki.1001-3563.2025.19.027

PDF(1877 KB)

Packaging Engineering ›› 2025, Vol. 46 ›› Issue (19) : 258-267. DOI: 10.19554/j.cnki.1001-3563.2025.19.027

Automatic and Intelligent Technology

Construction and Application of Precise Prediction Model for Cigarette Weight of Cigarette Equipment

TANG Hanxi¹, ZHANG Sheng^1*, XU Ran¹, LYU Xiaobo¹, CHEN Yanmeng¹, WU Yongjin¹, CHEN Wentao²

Author information +

History +

Abstract

The cigarette weight directly affects the forming quality of tobacco packaging. The work aims to propose an intelligent prediction method to reduce waste rate and improve quality stability to address the time-lag issue in cigarette rod weight control of cigarette equipment. With PROTOS 2C cigarette equipment as the research subject, a Stacking ensemble learning model integrating five algorithms (XGBoost, LGBM, LSTM, TCN, and Transformer) was established. The model achieved the ahead cigarette rod weight trend prediction through real-time data analysis and prediction algorithms, and related factors were quantified in real-time online, guiding technical personnel to improve the dimension and accuracy of equipment adjustment. Experimental results demonstrated superior performance of the model across evaluation metrics including Mean Squared Error (MSE), scatter plots, residual histograms, and Kernel Density Estimation (KDE). The trend of cigarette rod weight change could be predicted 5 seconds in advance. The number of cigarettes rejected by microwave detection was decreased by 86.70% compared with before, and the adjust time was decreased by 79.29% compared with before. Both stability and real-time performance of cigarette weight control were significantly improved. The Stacking ensemble model effectively resolves the time-lag issues inherent in traditional control methods, helps production to adjust equipment parts and parameters in advance, optimize production process, and improve the stability and consistency of cigarette quality.

Key words

PROTOS 2C cigarette equipment / cigarette rod weight / precise prediction / algorithm model / construction / weight control

Cite this article

EndNote

Ris (Procite)

Bibtex

Download Citations

TANG Hanxi, ZHANG Sheng, XU Ran, LYU Xiaobo, CHEN Yanmeng, WU Yongjin, CHEN Wentao. Construction and Application of Precise Prediction Model for Cigarette Weight of Cigarette Equipment[J]. Packaging Engineering. 2025, 46(19): 258-267 https://doi.org/10.19554/j.cnki.1001-3563.2025.19.027

References

[1] 于飞, 罗辉林, 柯凯, 等. 基于Stacking算法的转炉吹炼终点钢水磷含量预测[J]. 炼钢, 2024, 40(4): 11-16.
YU F, LUO H L, KE K, et al.Prediction of Phosphorus Content in Molten Steel at the End of Converter Blowing Based on Stacking Algorithm[J]. Steelmaking, 2024, 40(4): 11-16.
[2] KATE P, RAVI V, GANGWAR A.FinGAN: Chaoticgenerative Adversarial Network for Analytical Customer Relationship Management in Banking and Insurance[J]. Neural Computing and Applications, 2023, 35(8): 6015-6028.
[3] WANG Y C, HU W H, XIAO B J, et al.Fault Detection of Line-Averaged Plasma Density on EAST Using LightGBM[J]. Fusion Engineering and Design, 2025, 211: 114772.
[4] WEI J H, MIYAZAKI Y, KITAMURA K, et al.Construction of Collision-Type Prediction Model Based on Light Gradient Boosting Machine Algorithm and Elucidation of Data Causality and Relevance by Bayesian Network[J]. Traffic Injury Prevention, 2022, 23(sup1): S181-S182.
[5] 李自娟, 李宜馨, 赵海洋, 等. 基于机器学习的不同环境条件下松散回潮水分变化的预测[J]. 包装工程, 2024, 45(17): 119-128.
LI Z J, LI Y X, ZHAO H Y, et al.Prediction of Moisture Changes in Loosening and Conditioning Based on Machine Learning under Different Environmental Conditions[J]. Packaging Engineering, 2024, 45(17): 119-128.
[6] 王晓慧, 田天弘, 覃京燕, 等. 工业数字孪生数据建模在钢铁行业中的应用研究[J]. 包装工程, 2024, 45(8): 11-20.
WANG X H, TIAN T H, QIN J Y, et al.Application of Industrial Digital Twin Data Modeling in Iron and Steel Industry[J]. Packaging Engineering, 2024, 45(8): 11-20.
[7] 李壮年, 储满生, 柳政根, 等. 基于机器学习和遗传算法的高炉参数预测与优化[J]. 东北大学学报(自然科学版), 2020, 41(9): 1262-1267.
LI Z N, CHU M S, LIU Z G, et al.Prediction and Optimization of Blast Furnace Parameters Based on Machine Learning and Genetic Algorithm[J]. Journal of Northeastern University (Natural Science), 2020, 41(9): 1262-1267.
[8] 刘晓, 周荣喜, 李玉茹. 基于Stacking算法集成的我国信用债违约预测[J]. 运筹与管理, 2023, 32(3): 163-170.
LIU X, ZHOU R X, LI Y R.Default Prediction of Credit Bond in China Based on Stacking Algorithm Integrated Model[J]. Operations Research and Management Science, 2023, 32(3): 163-170.
[9] 田冬梅, 刘家鹏, 张芹, 等. 基于Stacking的集成算法在证券趋势预测中的应用[J]. 现代电子技术, 2022, 45(19): 115-121.
TIAN D M, LIU J P, ZHANG Q, et al.Security Movement Trend Prediction Based on Stacking Ensemble Algorithm[J]. Modern Electronics Technique, 2022, 45(19): 115-121.
[10] 杨瑞, 张海清, 李代伟, 等. 基于多尺度门控卷积与深度注意力的时序分类方法[J]. 软件导刊, 2025, 24(2): 33-39.
YANG R, ZHANG H Q, LI D W, et al.A Temporal Classification Method Based on Multi-Scale Gated Convolution and Depth-Wise Attention[J]. Software Guide, 2025, 24(2): 33-39.
[11] 范艺扬, 张洋, 曾尚, 等. 基于分解和频域特征提取的多变量长时间序列预测模型[J]. 计算机应用, 2024, 44(11): 3442-3448.
FAN Y Y, ZHANG Y, ZENG S, et al.Multivariate Long-Term Series Forecasting Model Based on Decomposition and Frequency Domain Feature Extraction[J]. Journal of Computer Applications, 2024, 44(11): 3442-3448.
[12] 袁战军, 王瑾, 谢利理. 基于DSP的新型烟支重量控制系统[J]. 海南大学学报(自然科学版), 2016, 34(4): 343-349.
YUAN Z J, WANG J, XIE L L.New Cigarette Weight Control System Based on DSP[J]. Natural Science Journal of Hainan University, 2016, 34(4): 343-349.
[13] 兰志超, 鄢红章, 张艮水, 等. 基于模糊算法的卷接机烟支重量自动控制系统优化研究[J]. 机械与电子, 2023, 41(1): 35-38.
LAN Z C, YAN H Z, ZHANG G S, et al.Research on Optimization of Cigarette Weight Automatic Control System of Coiler Based on Fuzzy Algorithm[J]. Machinery & Electronics, 2023, 41(1): 35-38.
[14] 段兴坤, 谢勇, 邓加睿, 等. 烟支单支克重偏差影响因素分析[J]. 价值工程, 2024, 43(9): 38-40.
DUAN X K, XIE Y, DENG J R, et al.Analysis of Influencing Factors of Weight Deviation of Single Cigarette[J]. Value Engineering, 2024, 43(9): 38-40.
[15] 方维岚, 陆正卿, 张晓杰. 基于斯皮尔曼相关性系数分析的烟支重量自动控制系统优化研究[J]. 电子元器件与信息技术, 2024, 8(10): 268-270.
FANG W L, LU Z Q, ZHANG X J.Research on Optimization of Automatic Control System of Cigarette Weight Based on Spearman Correlation Coefficient Analysis[J]. Electronic Components and Information Technology, 2024, 8(10): 268-270.
[16] 傅雅歆, 李树恩, 张晟, 等. 用神经网络实现烟支生产过程中的重量预测[J]. 上海轻工业, 2024(5): 168-170.
FU Y X, LI S E, ZHANG S, et al.Weight Prediction in Cigarette Production Process by Neural Network[J]. Shanghai Light Industry, 2024(5): 168-170.
[17] 周波, 张开桓, 吴芳基. 基于GPR模型的烟支重量控制系统启动位置预测方法及系统: 中国, 108323797A[P].2018-07-27.
ZHOU B, ZHANG K H, WU F J. Cigarette Weight Control System Starting Position Predicting Method and System Based on GPR Model: China, 108323797A[P].2018-07-27.
[18] 马春辉, 侯媛媛, 杨杰, 等. 基于RUN-XGBoost算法的土石坝渗流预测模型[J]. 水利水电科技进展, 2024, 44(2): 72-78.
MA C H, HOU Y Y, YANG J, et al.Seepage Prediction Model of Earth-Rockfill Dams Based on RUN-XGBoost Algorithm[J]. Advances in Science and Technology of Water Resources, 2024, 44(2): 72-78.
[19] 王红斌, 李志文. 基于CNN-LSTM混合神经网络的卷烟滤棒质量预测[J]. 长春理工大学学报(自然科学版), 2024, 47(4): 67-73.
WANG H B, LI Z W.Cigarette Filter Rod Quality Prediction Based on CNN-LSTM Hybrid Neural Network[J]. Journal of Changchun University of Science and Technology (Natural Science Edition), 2024, 47(4): 67-73.