基于自适应随机优化的连续阵风关键载荷预测

连续阵风载荷是构成民用飞机设计工况的主要载荷之一，在设计阶段，任意一轮的模型更新都涉及到上万种载荷工况的计算，然而其中仅个别工况构成载荷包线，需进行强度校核。为此发展了一套阵风关键载荷的快速识别方法。首先，采用二水平全因子（2LFF）采样获取得到初始计算工况，基于已计算得到的载荷值，结合多元自适应回归样条（MARS）建立一个可靠的代理模型；然后，在此基础上，开创性地应用自适应随机优化技术，实现对阵风关键工况及载荷的主动搜索；最后，以适航条款规定的侧向连续阵风载荷进行方法验证及参数影响研究。计算结果表明，本文建立的方法可以高效且准确地实现连续阵风关键载荷的预测，针对本文算例，关键载荷的预测值与基准值相比误差小于1%。

Prediction of critical continuous gust load based on adaptive stochastic optimization

Continuous gust load is one of the main loads for commercial airplane. In the design stage, any loop of model update involves the calculation of thousands of gust load cases; however, only a few cases constitute the critical load envelop, and are needed for strength check. Therefore, a rapid identification method for gust critical load is developed in this paper. First of all, using the two-level full factorial (2LFF) sampling technique, a set of initial calculation cases is obtained. With the calculated load values, a reliable surrogate model is established based on the multivariate adaptive regression spline (MARS). Secondly, on this basis, adaptive stochastic optimization technology is adopted to achieve active search of the critical gust case and load. Finally, using the test case of the lateral continuous gust load, the validation and the investigation on the influence of parameters are carried out. The calculation results show that the proposed method can effectively and accurately predict the gust critical load, and the prediction error of critical load is less than 1%.