A new method for singularity detection based on singular value decomposition

作者：Zeng, Zuo-Qin; Zhao, Xue-Zhi

来源：Journal of Shenyang University of Technology, 2011, 33(1): 102-107.

• Lipschitz
• Milling force
• Signal decomposition
• Singularity detection
• Vanishing moment

Aiming at the signal singularity detection problem, a construction method for the matrix with smaller column number and larger row number using one-dimensional signal series under the continuous signal intercepting way was proposed, and the detection was realized by means of singular value decomposition. The signal decomposition principle based on singular value decomposition under this matrix was analyzed. The singularity detection effect of this method was studied, and compared with the detection effect of Hankel matrix and wavelet transform. The proposed method was applied for detecting the singularity in a milling force signal. The experimental result shows that the small impact phenomenon caused by milling cutter or workpiece problems during the milling process can be effectively revealed by this method. The impulses indicating the position of singular points in all component signals have large amplitude and narrow width, and form the vivid contrast with the surrounding high-frequency noises, which is favorable to judge the position of singular points more accurately.

• 出版日期2011-2
• 单位华南理工大学