Title | ||
---|---|---|
Multi-frequency signal modeling using empirical mode decomposition and PCA with application to mill load estimation. |
Abstract | ||
---|---|---|
Multi-frequency signals consist of different time-scale components which have different physical interpretations. Normal principal component analysis (PCA) methods and frequency spectrum feature selection techniques do not work well in a multi-scale domain. This paper combines empirical mode decomposition (EMD), PCA, and an optimal feature extraction method to extract, select and model different scale frequency signals. We successfully apply this approach to a laboratory scale wet ball mill. The shell vibration signal produced by the ball mill of the grinding process is used for modeling the mill load. The experimental results demonstrate that this novel approach is effective compared with the other existing methods. |
Year | DOI | Venue |
---|---|---|
2015 | 10.1016/j.neucom.2014.08.087 | Neurocomputing |
Keywords | Field | DocType |
Feature selection and extraction,Empirical mode decomposition,Multi-frequency signals,Mill load estimation | Mill,Ball mill,Signal modeling,Feature selection,Pattern recognition,Feature extraction,Artificial intelligence,Vibration,Machine learning,Principal component analysis,Mathematics,Hilbert–Huang transform | Journal |
Volume | ISSN | Citations |
169 | 0925-2312 | 5 |
PageRank | References | Authors |
0.41 | 6 | 4 |
Name | Order | Citations | PageRank |
---|---|---|---|
zhuo liu | 1 | 7 | 2.52 |
Tianyou Chai | 2 | 2014 | 175.55 |
wen yu | 3 | 9 | 2.28 |
Jian Tang | 4 | 526 | 148.30 |