Tool-wear monitoring during micro-end milling using wavelet packet transform and Fisher’s linear discriminant

摘要: Tool wear is one of the most important parameters in micro-end milling, and can be used to monitor condition machine tool. A mill has different characteristics from a macro-scale end mill; particular, shank run-out (which negligible tool due low aspect ratio) significant inducing excessive reduced life leading sudden, premature failure. In this paper, novel tool-wear monitoring method described for determining state using wavelet packet transforms Fisher’s linear discriminant. Force torque signals were measured dynamometer reflect geometric changes wear. Because small signal-to-noise ratio, sensor during milling process periodically averaged, resulting single-period provided improved efficiency feature extraction transforms. The extracted features classified domain determine employing hidden Markov model. recognition results compared with those an energy-based technique, we found that our could more accurately both normal failure mills.