Abstract | ||
---|---|---|
The data obtained from any process may be corrupted with noise and outliers which may lead to false-alarm when applying conventional PCA to process monitoring. To overcome the above mentioned limitations of conventional PCA, an approach is developed by combining the ability of wavelets and moving median filter with PCA. This method utilizes the quality of wavelets and moving median filter to preprocess the data to eliminate noise and outliers. At last, this method is applied to fault detection and has a good effect which proves the method is effective and feasible. |
Year | DOI | Venue |
---|---|---|
2007 | 10.1007/978-3-540-72393-6_118 | ISNN (2) |
Keywords | Field | DocType |
median filter,good effect,component analysis,fault detection,conventional pca,hybrid principal,principal component analysis | Control limits,Median filter,Pattern recognition,Fault detection and isolation,Computer science,Outlier,Mean squared error,Artificial intelligence,Principal component analysis,Machine learning,Wavelet | Conference |
Volume | ISSN | Citations |
4492 | 0302-9743 | 0 |
PageRank | References | Authors |
0.34 | 2 | 3 |
Name | Order | Citations | PageRank |
---|---|---|---|
Chenglin Wen | 1 | 179 | 42.72 |
Shao-Hui Fan | 2 | 0 | 0.34 |
Chen Zhi-Guo | 3 | 1 | 2.04 |