文档介绍：: .
推进技术
异常检测和修复。在甲烷预混对冲火焰数据集
上，将异常划分为两种类型，并搭建 U-Net 卷积神经网络架构。经过训练和优化，模型以较
高置信水平识别两类异常并使用不同策略自适应修复，过滤噪声并保留原始正常数据。同时
模型具有较好的可迁移性，可以为其它种类的流场数据修复提供参考。与 POD 迭代法和中值
滤波相比，神经网络强大的非线性特征具有明显的优势，这种方法不仅修复率高，而且在不
同工况下鲁棒性好。
关键词：PIV；后处理；U-Net；异常检测；异常修复
中图分类号： 文献标识码：A
Inpainting PIV Flow Fields with Deep Learning
QIN Zi-yu1, ZHENG Dong-sheng1, ZHOU Yu-chen1, HAN Xiao2*, HUI Xin2, WANG Zuo-xia1,
LIU Xiang1, ZHANG Chi2
(1. National Key Laboratory of Science and Technology on Aero-Engine Aero-thermodynamics，
School of Energy and Power Engineering，Beihang University，Beijing 100191，China;
2. National Key Laboratory of Science and Technology on Aero-Engine Aero-thermodynamics,
Research Institute of Aero-Engine, Beihang University, Beijing, 100191, P. R. China)
Abstract: Particle Image Velocimetry(PIV) is widely used to measure the flow fields in
aerospace researches. However, it can hardly tackle complex flow fields in combustors,
and defects could usually be found in the flow field after the post-processing, where
traditional cross-correlation method is applied. The deep learning method is applied in
the PIV post-processing to achieve detection and inpainting of abnormal flow field data.
On the counterflow premixed methane flame dataset, the abnormal data is classified
into two categories, which is fixed with a U-Net-based convolutional neural network
model. After training and optimizing, the model can detect anomalies at