A Method for Searching for Optimal Parameters of the Technological Process of Impregnation of Three-Dimensional Reinforced Composite Products Based on Machine Learning Models

A new approach to developing an adaptive manufacturing process for composite materials is presented. An experimental sample created using 3D weaving is studied, and a rapid engineering model is created to optimize the parameters of the preform impregnation process. The model uses data on the preform and resin parameters to determine the quality characteristics of the final part. According to the results obtained, the proposed method significantly reduces the number of surface defects in samples and allows for predicting injection time, which is critical for optimizing the manufacturing process. The conducted studies confirmed the high efficiency of the developed model and substantiated its practical application in industry.