| 基于一维卷积神经网络的发动机气路故障诊断研究 |
投稿时间：2020-08-25 修订日期：2020-09-17  点此下载全文 |
| 引用本文： |
| 摘要点击次数: 144 |
| 全文下载次数: 0 |
|
| 基金项目:国家自然科学基金项目（面上项目，重点项目，重大项目） |
|
| 中文摘要:气路故障诊断是航空发动机预测与健康管理的重要组成部分。通过将气路诊断信息转化为维修操作建议，可以方便工程人员确定发动机需要更换和维修的部件范围，降低维修成本。结合深度学习可以对数据特征进行自动获取的特点，提出了一种基于一维卷积神经网络的气路故障诊断方法；训练的诊断模型在测试集上的准确率达到了98.4%，有较好的诊断效果。 |
| 中文关键词:航空发动机，气路故障，一维卷积神经网络，GSP |
| |
| Engine Gas Path Fault Diagnostic Research Based on One-Dimensional Convolution Neural Network |
|
|
| Abstract:Gas path analysis is an important part of engine prognostics and health management.By converting the gas path diagnosis information into maintenance operation suggestions, it is convenient for engineers to determine the range of parts that need to be replaced and repaired, and the maintenance cost is reduced.Combining the characteristics of deep learning that can automatically obtain data features, this article proposes a gas path fault diagnosis method based on one-dimensional convolution neural network. Trained a new network model, with the accuracy of 98.4%, experimental results show that the network has good diagnostic effect. |
| keywords:aero-engine gas path fault one-dimensional convolution neural network GSP |
| 查看全文 查看/发表评论 下载pdf阅读器 |
位访问者,今日访问
次