基于分维LLE和Fisher判别的故障诊断方法
Fault diagnosis approach based on fractal dimension LLE and Fisher discriminant
查看参考文献15篇
文摘
|
针对非线性系统故障诊断难以解决的问题,通过改进的局部线性嵌入映射算法解决了非线性数据的特征映射问题.首先,通过线性拟合改进了基于分形维估计的内在维数的估计.然后,将故障状态与空间分布结合起来,通过确定数据点在空间超球内的分布完成故障的检测,在这个过程中将超球的确定与LLE算法中基于核函数的样本外数据扩展结合起来,大大减少了计算量,提高了算法的实时性.然后,利用Fisher判别分析进行故障匹配,通过计算最优的投影向量与历史故障数据投影向量的相似度的计算,完成故障识别,从而为复杂非线性系统故障诊断提供了一种新的有效的方法. |
其他语种文摘
|
Aiming at the difficult problem of nonlinear system fault diagnosis, an improved method based on LLE is used to solve the problem of character projection of nonlinear data. Firstly, the method of intrinsic dimension estimation based on fractal dimension is improved by linear fitting. And then sensor fault state is combined with spatial distribution to complete the fault detection. Out of sample extension is also considered and combined with fault detection algorithm, which can reduce computation task obviously and improve real time capability of the algorithm. Then, Fisher discriminant is used for fault type matching by computing the similarity between the optimal projection vectors and historical projection vectors, and fault diagnosis is completed. The proposed algorithm provides an effective method for nonlinear sensor fault diagnosis. |
来源
|
仪器仪表学报
,2010,31(2):325-333 【核心库】
|
关键词
|
局部线性嵌入(LLE)
;
故障诊断
;
非线性降维
;
内在维数
;
Fisher判别
|
地址
|
中国科学院沈阳自动化研究所, 机器人学国家重点实验室, 辽宁, 沈阳, 110016
|
语种
|
中文 |
文献类型
|
研究性论文 |
ISSN
|
0254-3087 |
学科
|
自动化技术、计算机技术 |
基金
|
国家863计划
|
文献收藏号
|
CSCD:3821345
|
参考文献 共
15
共1页
|
1.
徐宝国. 基于小波包变换和聚类分析的脑电信号.
仪器仪表学报,2009,30(1):24-28
|
被引
1
次
|
|
|
|
2.
杨进. 基于相关分析和近似熵的管道泄漏声信号特征提取及辨识方法.
仪器仪表学报,2009,30(2):272-279
|
被引
31
次
|
|
|
|
3.
SAMUEL L. Some studies in machine learning using the game of checkers.Part II.
IBMJournalResearch and Development,1967,11(4):601-608
|
被引
1
次
|
|
|
|
4.
桑海峰. 基于二维主成分分析的掌纹识别研究.
仪器仪表学报,2008,29(9):1929-1933
|
被引
19
次
|
|
|
|
5.
DUNIA R. Use of principal component analysis for sensor fault identification.
Computers chem.Eng,1996(l.20):713-718
|
被引
1
次
|
|
|
|
6.
JACKSON J E. Principal component and factor analysis:Part I-principal component.
JOURNAL OF QUALITY TECHNOLOGY,1980(12):201-21
|
被引
1
次
|
|
|
|
7.
HASTIE T.
Principle curves and surfaces,1984
|
被引
1
次
|
|
|
|
8.
ROWEIS S. Nonlinear Dimensionality Reduction by Locally Linear Embedding.
SCIENCE,2000,290(22):2323-2326
|
被引
1300
次
|
|
|
|
9.
DONOHO D L. Hessian eigenmaps:locally linear embedding,techniques for high-dimensional data.
Proceedings of the National Academy of Sciences,2003:5591-5596
|
被引
96
次
|
|
|
|
10.
DE RIDDER D.
Supervised locally linear embedding,2003:333-341
|
被引
2
次
|
|
|
|
11.
KEGL B. Intrinsic dimension estimation using packing numbers.
Neural Information Processing Systems 15 (NIPS'2002),2003
|
被引
1
次
|
|
|
|
12.
GRASSBERGEL J. Measuring the strangeness of strange attractor.
Physical D,1983(9):189-208
|
被引
1
次
|
|
|
|
13.
DOWNS J J. Plant-wide industrial process control problem.
Computers & Chemical Engineering,1993,17(3):245-255
|
被引
295
次
|
|
|
|
14.
BENGIO Y.
Out-of-sample extensions for LlE,isomap,MDS,Eigenmaps,and spectral clustering,2003
|
被引
2
次
|
|
|
|
15.
TAYLOR J S.
Kernel methods for pattern analysis,2004
|
被引
15
次
|
|
|
|
|