Method of Fault Pattern Recognition Based on Laplacian Eigenmaps
XU Fei-yun · Jisuanji fangzhen · 2008
The Laplacian Eigenmaps algorithm can effectively extract the low dimension feature embedded in high dimension nonlinear data. It was introduced into the fault diagnosis field and was applied for fault pattern recognition problem,and a new method of fault pattern recognition based on the Laplacian Eigenmaps (LE-FPR) was proposed. A nonlinear dimensionality reduction algorithm based on Laplacian Eigenmaps was used to directly learn original fault signal and extract the intrinsic manifold feature in data set. The method greatly preserved the whole geometry structure information embedded into the signal,overcame the flaw of conventional pattern recognition methods which only obtained the local linear structure in data set,and obviously improved the classification performance of fault pattern recognition. The simulation and instance results demonstrate the feasibility and effectiveness of the new method.