Deep learning-based intelligent control of moisture at the exit of blade charging process in cigarette production

• DOI: 10.2478/amns-2024-0026
• Publication Date: 2024-02-13T15:40:43Z
• Authors: Jinsheng Rui; Dongchen Qiu; Shicong Hou; Jing Rong; Xiaoxiao Qin; Jianan Fan; Kai Wu; Guoliang Zhao; Chengwen Zhu
• Keywords: 0202 electrical engineering, electronic engineering, information engineering; 02 engineering and technology

Abstract Currently, in the production of cigarettes blade, charging export moisture control means is relatively single and can not effectively guarantee excellent quality cigarette filament. In this paper, first all, working principle tobacco blade machine introduced, leaf for dynamically analyzed, influence return air temperature on process explored. Secondly, based traditional PID controller, an adaptive fuzzy controller established by combining rules, then stacked noise-reducing self-encoder deep learning combined with to design intelligent structure process. Finally, effectiveness intelligence control, capability index, effect before after application were analyzed controlled experiments, respectively. The results show that difference between predicted value real paper’s method only 0.5%, index improved 1.48 compared it range 56.86℃~57.21℃. introduced accurately dosing process, which ensures filament making.