| 基于最小邻近准则的故障类型识别方法 |
| |
| 引用本文： | 李登元,赵述芳,谢桂海,高海峰.基于最小邻近准则的故障类型识别方法[J].军械工程学院学报,1999(3). |
| |
| 作者姓名： | 李登元 赵述芳 谢桂海 高海峰 |
| |
| 作者单位： | 军械工程学院研究生队!石家庄，050003，河北省电子工业学校!石家庄，050061，军械工程学院计算机教研室!石家庄，050003，军械工程学院研究生队!石家庄，050003 |
| |
| 摘 要： | 对设备的多故障类型自动识别问题是设备进行自动诊断的难题之一。讨论基于信号频域特性的故障信号特征提取方法，利用这种方法可以把统计学的聚类分析技术用于故障类型识别，建立起一个由信号特征向量构成的向量空间，以最小邻近准则作为故障类型识别的依据。
|
| 关 键 词： | 分类识别 最小邻近准则 特征提取 |
INVESTIGATION OF FAULT IDENTIFICATION METHOD BASED ON THE NEAREST NEIGHBOURHOOD CRITERIA |
| |
| Li Dengyuan, Zhao Shufang, Xie Guihai, Gao Haifeng.INVESTIGATION OF FAULT IDENTIFICATION METHOD BASED ON THE NEAREST NEIGHBOURHOOD CRITERIA[J].Journal of Ordnance Engineering College,1999(3). |
| |
| Authors: | Li Dengyuan Zhao Shufang Xie Guihai Gao Haifeng |
| |
| Abstract: | The automatic identification and classification of multi - fault of device is one of the most difficult topics in the automation diagnosis of device. A feature extraction method in frequency domain of signal is proposed in this paper. The cluster analysis algorithm in mathematical statistics can be used in fault identification by this method. A metric space of signal feature vector is established. In terms of the nearest neighbourhood principle, the rules of fault identification and classification can be derived. The criteria can be considered as the reasoning deduction of expert system for fault automation identification. |
| |
| Keywords: | classfication criteria of nearest neighbourhood feature extraction |
| 本文献已被 CNKI 等数据库收录！ |
