基于代理模型的高超声速飞行器外形参数优化

针对高超声速飞行器分析复杂且难度较大,提出了一种代理模型的构建方法,使用代理模型近似替代性能分析与优化过程中含有复杂学科耦合的机理模型。根据巡航任务需求,确定了优化目标为静动态性能最优与模型差异最小。使用灵敏度分析的方法,建立了代理模型。将代理模型进行静动态性能分析,并与机理模型配平结果进行了对比验证,发现两个模型的配平特性趋势是完全一致的,迎角的数值差不足3%,升降舵偏转角的数值差仅在前体下倾角较大时偏大,约为20%。基于构建的代理模型与优化的性能指标,对模型的外形参数进行了配平性能优化与间隙度量优化,并与机理模型的优化结果与优化效率进行对比,发现两者结果相差不足2%,但使用代理模型的优化效率提高了456%,证明了基于代理模型的优化可以在确保精度的基础上提高优化效率。 

Shape parameters optimization of hypersonic vehicle based on surrogate model

In view of the complexity and difficulty of hypersonic vehicle analysis, a surrogate model was proposed to approximate the true model with complex disciplinary coupling in the process of performance analysis and optimization. According to the requirements of cruise mission, the objective of optimization was to optimize the static and dynamic performance and minimize the model difference. By means of sensitivity analysis, the surrogate model was established. The static and dynamic performance analysis of the surrogate model was carried out, and the results of the true model were compared to verify the accuracy of the surrogate model. It was found that the trend of the trimming characteristics of the two models was completely consistent, the difference of angle of attack was less than 3%, and the difference of elevator deflection angle was only about 20% when the lower angle of the front body was large. Based on the surrogate model and optimized performance index, the trimming performance optimization and gap metric optimization of the surrogate model’s shape parameters were compared with the optimization results of the true model and the optimization efficiency. It was found that the difference between the two results was less than 2%, but the optimization efficiency using surrogate model was improved by 456%. It was proved that the optimization based on the surrogate model can improve the optimization efficiency without compromise of the accuracy.