基于机器学习的进口导流叶片损失与落后角预测

Inlet Guide Vane Loss and Deviation Prediction Based on Machine Learning

准确预测进口导流叶片的损失与落后角对设计优化压气机性能,提升压气机稳定性至关重要。本文使用叶型设计程序与MISES求解器耦合成的计算模型,创建了进口导流叶片样本数据库。通过支持向量回归搭建了损失及落后角代理模型,对比代理模型预测结果与经验模型计算结果,发现此代理模型能准确预测正反预旋大攻角下进口导流叶片的性能。

Predicting the loss and deviation of the Inlet Guide Vanes’(IGV)is important to improve the aerodynamic performance and stability of the compressors.In this paper,an analysis model consists of profile design program and MISES is conducted to create the IGV database.By comparing the results from the empirical model and the surrogate model built by support vector regression,it is found that the surrogate model can accurately predict the aerodynamic performance of the IGV in the environment of positive pre-swirl and negative pre-swirl at large incidence.