基于Isight的压气机三维叶片鲁棒性优化

打开文本图片集

摘 要：以Isight软件为基础，搭建叶片鲁棒性优化平台，基于构建的RBF神经网络模型和蒙特卡罗方法对压气机Rotor37进行鲁棒性优化。优化结果表明：压气机性能曲线整体向左上方移动，裕度几乎不变，选取的两个优化工况处效率的均值分别提高了0.24%和0.46%，方差分别降低了16.3%和15%。

关键词：压气机；叶片；鲁棒性优化；RBF神经网络；蒙特卡罗方法；效率；裕度

中图分类号：TH138.5  文献标志码：B  文章编号：1671-5276（2024）05-0126-04

Robustness Optimization of Compressor Three-dimensional Blades Based on Isight

Abstract：A blade robustness optimization platform was built based on Isight software， and with the constructed RBF neural network model and by Monte Carlo method， the robustness optimization of Rotor37 was carried out. The optimization results show that the overall performance curve moves to the left and up with nearly no change of margin， the average efficiency increases by 0.24% and 0.46%， and the variance decreases by 16.3% and 15% respectively at the two selected optimization conditions.

Keywords：compressor；blade；robustness optimization；RBF neural network；Monte Carlo method；efficiency；margin

0 引言

叶片是组成航空发动机的重要部件，但在加工过程中会不可避免地出现加工误差。（剩余4719字）