Analysis of Dynamic Mechanical Properties of Glass Fiber Fabric Composites Based on Deep Learning

A deep learning-based performance characterization and optimization method is proposed for the dynamic mechanical properties of glass fiber fabric composites. By combining the multi-factor interaction model (MFIM) and Genoa-MCQ optimization algorithm, the mechanical behaviors of glass fiber and epoxy resin matrix composites are analyzed, and virtual testing and iterative optimization are performed. The deep learning model was used to process the multi-level physical data, and the performance prediction and optimization from micro to macro were successfully achieved, especially the tensile, compressive and shear strengths were significantly improved. The optimization results provide a more accurate prediction of the mechanical properties of the materials, further validating the feasibility and effectiveness of virtual testing in composites research. This method provides new ideas and technical support for the performance evaluation and design of composite materials.