Research on Powder Formulation Design Based on Machine Learning

In this study, the machine learning technology was used to establish the propellant performance prediction model, and the propellant force, explosion temperature and explosion heat were predicted respectively. Then, based on the established propellant performance prediction model, combined with evolutionary algorithm, a propellant formulation optimization method with energy as the goal is proposed. Compared with the standard propellant M30, the detonation temperature of the optimized propellant formulation is reduced by 100K without reducing the propellant force. This study provides a new idea for the design of propellant formula, and is expected to improve the efficiency of propellant formula design.