Digital Twin and Edge Computing-Driven MES Intelligent Decision-Making in Tobacco Enterprises

With the advancement of intelligent manufacturing technologies, manufacturing execution system (MES) as the core bridge connecting enterprise management and production workshop, its data processing capability directly affects the operational efficiency of enterprises. Tobacco industry as a typical representative of fine chemical and discrete manufacturing, its production process is complex and variable, the real-time production data, accuracy and completeness of the higher requirements, but the current MES system for tobacco companies generally has serious data processing delays, information integration difficulties, decision-making response lag and other issues. To address these challenges, this paper proposes a basic data model of MES based on the fusion of digital twin and edge computing, which realizes data processing and filtering in the vicinity of the production site by deploying edge computing nodes and reduces the burden of the central server. At the same time, we construct a digital twin model with multi-dimensional interactions of equipment, material, process and quality, realize the real-time mapping and two-way feedback between the physical world and the digital space, and design an adaptive intelligent decision-making algorithm, which is able to automatically adjust the process parameters according to the changes of the production state. A six-month comparative test in tobacco enterprise A shows that the model improves data processing efficiency by 75.3%, reduces system response time to less than 35 ms, improves production plan execution accuracy by 28.4%, and reduces the fluctuation range of key quality indicators by 45.2%, which is especially effective in the control of cigarette density and detection of splicing quality anomalies. The research results not only provide practical solutions for intelligent manufacturing in the tobacco industry, but also provide a technical path for the digital transformation of other manufacturing industries with similar production characteristics and help to promote the manufacturing industry from the traditional passive response mode to the predictive, intelligent production management mode of change.