Drill wear monitoring by using hidden Markov model based on wavelet decomposition of power spectrum

In drilling process, the power spectrums of drilling force have been proved tightly related to the tool wear, and which are widely applied in the monitoring of tool wear. But the extraction and identification of the features of power spectrum have always been an unresolved difficult problem. This paper achieves it through decomposition the power spectrum in multilayer using wavelet transform and extraction the low frequency decomposition coefficient as the envelope information of the power spectrum. Meanwhile, with an aim at the strong randomization and uncertainty characteristic between the feature vectors and drill wears in drilling process, a kind of drill wear monitoring method based on Hidden Markov Model (HMM) is presented. The experimental results indicate that the features of power spectrum can be extracted efficiently by using wavelet transform, and that the statistics model for the low frequency decomposition coefficient of the power spectrum in each drill wear condition can be established using HMM which can effectively track the developing trends of drill wears so as to realize the monitoring of drill wear states and service life.