Mathematical Models and Data Processing Methods for Improving the Efficiency of High-Tech Production of Polymeric Films

Mathematical models and data processing methods used for the digitalization of high-tech production of polymeric films, characterized by flexibility, large-capacity, multi-assortment, returnable waste, have been developed. The models describe the physical processes at the main stages of production (preparation of the extrudate, forming the extrudate, fixing the film structure), take into account the variability of equipment configurations, and make it possible to calculate the quality indices of the extrudate and film that are not monitored in production. Machine learning methods are used to process production data with large volume, high accumulation rate, variety of sources, and predict the quality of films. Mathematical models and methods have been implemented in the form of a software package that is adjusted for various types of films and configurations of production systems. It helps to solve the tasks of controlling the consumer characteristics of films (thickness difference, number of surface defects of various types, shrinkage, color, telescoping). Testing according to the extrusion-calender production of pharmaceutical and food packaging films at factories in Russia and Germany has confirmed the adequacy of the models and the operability of the software package. Their use contributes to improving the production efficiency by increasing the yield of high-quality film with the same amount of raw materials and the possibility of increasing film price for customers due to a significant improvement in the consumer characteristics.