Multiobjective Optimization Method of Solid Rocket Motor Finocyl Grain Based on Surrogate Model

To improve the performance of a solid rocket motor (SRM), multiobjective optimal design method that can consider structural integrity, internal ballistic performance, and loading SRM was proposed based on parametric modeling surrogate technology. Firstly, technology introduced into field integrity analysis for high-loading SRM, which influences load geometric parameters maximum von Mises strain grain were analyzed, effectively improved sampling speed prediction accuracy model. Combining calculation models combustion surface area volume fraction Pareto solution set obtained NSGA-II algorithm. Under constraints optimization model, be reduced by up to 26.72% increased 1.83% compared with original. In addition, in this paper is significantly superior efficiency, capable reducing both single time more than 95% number numerical simulations from 20,000 400, average deviation only 1.87%.