Optimization of precision molding process parameters of viscoelastic materials based on BP neural network improved by genetic algorithm

Abstract

Micro-structured optical components (MOC) are popular due to their unique optical and mechanical properties. Glass molding processing (GMP) is currently the main method of MOC manufacturing. In GMP, the process parameters determine the efficiency and quality of the glass processing. However, the optimization of their multi-objective molding process parameters has not been studied comprehensively enough. In this paper, a thermocompression model of a viscoelastic material and a mold is developed using coupled thermal-structural analysis in Abaqus . In addition, the data was analyzed to investigate the influence of process parameters on GMP performance. Then, a BP neural network model was built, and the model was trained and tested. A prediction model that can reasonably reveal the relationship between the process parameters and the molding effect was obtained. A multi-objective optimization algorithm GA was used to optimize the design of the hot-pressing process parameters. At last, a set of process parameters with the best forming effect was obtained. This study provides valuable insights into optimizing glass molding process parameters for practical production.