基于特征注意力机制和ResNet的航空发动机剩余寿命预测  

REMAINING USEFUL LIFE PREDICTION OF AERO-ENGINE BASED ON FEATURE ATTENTION MECHANISM AND RESNET

作　　者：李雅 孙行行 侯彦东[1] Li Ya;Sun Hanghang;Hou Yandong(School of Artificial Intelligence,Henan University,Zhengzhou 450000,Henan,China)

机构地区：[1]河南大学人工智能学院,河南郑州450000

出　　处：《计算机应用与软件》2025年第3期22-28,47,共8页Computer Applications and Software

基　　金：国家自然科学基金项目(61374134)。

摘　　要：在航空发动机的剩余寿命预测中,针对大多数深度学习方法没有对不同传感器输入特征进行自适应加权的问题,提出一种基于特征注意力机制和残差网络(ResNet)的剩余寿命预测模型(FAResNet)。使用带有压缩激励的特征注意力机制模块对多传感器输入特征进行自适应加权处理;使用两个并行的ResNet分支分别沿时间维度提取时序特征和沿传感器特征维度进行特征融合,深度挖掘数据的隐藏特征。在公开数据集上的仿真实验验证了所提出模型的有效性,与其他深度学习模型相比有着更低的预测误差。In the remaining useful life prediction of aero-engine,a remaining useful life prediction model(FAResNet)is proposed based on feature attention mechanism and residual network(ResNet)to solve the problem that most deep learning methods do not carry out adaptive weighting for different sensor input features.A feature attention mechanism module with squeeze excitation was used for adaptive weighting processing of multi-sensor input features.Two parallel ResNet branches were used to extract time series features along the time dimension and feature fusion along the sensor feature dimension respectively to deeply mine the hidden features of the data.Simulation experiments on open data sets demonstrate the effectiveness of the proposed model,which has lower prediction errors than other deep learning models.

关 键 词：剩余寿命 注意力机制 残差网络 涡扇发动机 

分 类 号：TP202.1[自动化与计算机技术—检测技术与自动化装置]