Design and Analysis of Optimal Recipe Prediction Model Based on Deep Learning for Advanced Composite Material Injection Molding

Manufacturing recipe management is essential for product quality and efficient production in the injection molding manufacturing process. If the recipe setting is not optimized, various problems such as the cause of equipment failure, decrease in production efficiency, and increase in defect rate may occur. The injection molding manufacturing recipe is influenced by a combination of factors of various manufacturing conditions. In particular, with the recent development of the electric vehicle market, differences from general injection molding technologies exist in manufacturing electric vehicle battery cases using injection molding technology. It includes technology that adds advanced composite materials to the surface of the battery case and to make products that are durable against high heat and pressure. Derivation of injection molding recipes using high-tech composite materials has more complex factors affecting product production than general injection molding, so it is difficult to expect products of uniform quality if recipes are set based on the empirical know-how of existing experts. The production efficiency is also not constant, causing problems. For the above reasons, this paper proposes a model that can derive an optimal recipe for injection molding recipes using advanced composite materials using the MLP algorithm.