首页 | 本学科首页   官方微博 | 高级检索  
     检索      

基于RDK-ELM 的航空发动机控制系统故障诊断
引用本文:陈虹潞,黄向华.基于RDK-ELM 的航空发动机控制系统故障诊断[J].航空发动机,2020,46(2):34-40.
作者姓名:陈虹潞  黄向华
作者单位:南京航空航天大学能源与动力学院,南京210016,南京航空航天大学能源与动力学院,南京210016
基金项目:国家自然科学基金(51576097)资助
摘    要:为保持较高诊断正确率,缩短训练时间,满足航空发动机故障诊断对于实时性和高诊断率的需求,提出1种对深度核极限学习机的简约改进方法。输入数据中随机选取部分数据作为支持向量,结合深度学习网络的多层结构,完成了对输入样本的特征提取,通过核函数实现了高维空间映射分类。数字仿真表明:算法分类正确率高,训练时间短,可应用于航空发动机控制系统的故障诊断。

关 键 词:控制系统  故障诊断  深度学习  极限学习机  简约改进  航空发动机

Fault Diagnosis of Aeroengine Control System Based on RDK-ELM
Authors:CHEN Hong-lu  HUANG Xiang-hua
Institution:(College of Energy and Power Engineering,Nanjing University of Aeronautics&Astronautics,Nanjing 210016,China)
Abstract:In order to maintain the high diagnostic accuracy and shorten the training time,a method of Reduced Deep Kernel Extreme Learning Machine(RDK-ELM)was proposed for meeting the needs of aeroengine fault diagnosis for real-time and high diagnostic rate.Partial data was selected randomly as support vector in the input data. The feature extraction of the input sample was accomplished by combining the multi-layer structure of the deep learning network. The high-dimensional spatial mapping classification was realized by kernel function. Digital simulation shows that the algorithm has high classification accuracy and short training time,and can be applied to fault diagnosis of aeroengine control system.
Keywords:control system  fault diagnosis  deep learning  extreme learning machine  reduced method  aeroengine
本文献已被 维普 万方数据 等数据库收录!
点击此处可从《航空发动机》浏览原始摘要信息
点击此处可从《航空发动机》下载免费的PDF全文
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号