Wavelet-based process quality monitoring and diagnosing

In order to improve machining quality, this paper presents an integrated process quality monitoring and diagnosing methodology for machining process. Dimensional quality characteristic of work-piece is first monitored by wavelet-based SPC charts. And then, power spectrum analysis is used to extract abnormal feature signal if abnormal change is detected. Basing on this, a correlated model is established to calculate correlation coefficient between abnormal change and its fault sources, and synthesised diagnosis is provided to determine the importance level of these fault sources. At last, a running example is given to verify the proposed methodology.