Fault Diagnosis Based on the Covariance Matrix Recursive
Yucheng Hu · Journal of Hangzhou Dianzi University · 2010
In the large-scale complex systems,local fault can propagates to some other components and cause serious consequences.Therefore the faults can be diagnosis timely.The conventional PCA method needs modeling off-line and detecting on-line.It has bad instantaneity.Therefore,this paper gives the covariance matrix recursive method that analysis the real-time or off-line data,and shows it is a kind of effective fault diagnosis method.At last,we compared the PCA with the covariance matrix recursive through an application in fault diagnosis,and find the covariance matrix recursive is effectual.