Journal of Engineering, Project, and Production Management, 2026, 16(6), 2026-125
Steady-State Control and Optimization of Process Quality of High-Speed Unit in Wrapping Workshop
1 Engineer,
Process Quality Department, Hebei Baisha Tobacco Co., Ltd, Shijiazhuang
052165, China, E-mail: shuxi@hpu123.uu.me
2 Director, Process Quality
Department, Hebei Baisha Tobacco Co., Ltd, Shijiazhuang 052165, China,
E-mail: Jinghui@hpu123.uu.me
3 Chief Inspector, Process Quality
Department, Hebei Baisha Tobacco Co., Ltd, Shijiazhuang 052165, China,
E-mail: zhijun@hpu123.uu.me
4 Senior Engineer, Silk Making
Workshop, Hebei Baisha Tobacco Co., Ltd, Shijiazhuang 052165, China,
E-mail: shaolong15@outlook.com (corresponding author).
Production Management
Received January 25, 2026; revised March 31, 2026; accepted June 1, 2026
Available online July 2, 2026
Abstract: High-speed equipment in the wrapping workshop exhibits strong nonlinearity, multi-variable coupling, and frequent disturbances in actual production. This makes it difficult for the traditional control method to maintain the stable consistency of tension, material supply and rolling pressure for a long time at high speed and is prone to steady-state deviation and quality fluctuations. In response to the above shortcomings, this study proposes a steady-state quality control method based on deep reinforcement learning. This method achieves adaptive adjustment of the high-speed wrapping process by constructing a dynamic prediction model, a policy network, a value evaluation network, and a steady-state deviation suppression mechanism. Experiments showed that in typical steady-state tests, the average steady-state deviation of tension, feed amount, and rolling pressure of the proposed method was only 0.04, which was 70% lower than that of the traditional control method. Under continuous disturbance conditions, the drift amplitude of the proposed variables was always controlled within 0.10. In the multi-source disturbance scenario, the Root Mean Square Error (RMSE) of the proposed method remained within a low range (0.15-0.23), maintaining high output consistency under strong disturbance conditions. The constructed intelligent control framework could automatically learn control rules without the need for precise mathematical models and achieve steady-state quality maintenance of high-speed wrapping equipment under complex disturbances. This study provides a feasible technical path for smart manufacturing and stabilization of cigarette production.
Keywords: High-speed wrapping process; steady-state quality control; deep reinforcement learning; multi-variable coupling; disturbance suppression; intelligent manufacturing.
Copyright © Journal of Engineering, Project, and Production Management (EPPM-Journal).
This work is licensed under a Creative Commons Attribution-NonCommercial-NoDerivs 3.0 Unported License.
Requests for reprints and permissions at eppm.journal@gmail.com.
Citation: Guo, S., Guo, J., Xu, Z., and Han, S. (2026). Steady-State Control and Optimization of Process Quality of High-Speed Unit in Wrapping Workshop. Journal of Engineering, Project, and Production Management, 16(6), 2026-125.
