Quality monitoring method based on tool vibrations and the discrete hidden markov model at various cutting parameters in hard turning

In turning operations, Surface roughness (Ra) and straightness (SRS) are important requirements for many workpieces. How to implement SRS monitoring is a crucial task. In this paper, a novel SRS monitoring method based on tool vibrations and the discrete hidden markov model (DHMM) is proposed. First, an in-depth theoretical analysis for the derived formulas of the SRS was conducted and found that the SRS are strongly associated with the vibration energy characteristic (VEC), so VEC was determined to serve as the characteristic for the SRS monitoring. Then, relying on the process quality criteria, SRS were divided into the limited workpiece quality intervals, which made the SRS monitoring becoming a kind of pattern recognition. Lastly, based on the DHMM, the identify and classify calculation process was provided. The experimental results have been found to agree well with the theoretical analyses and the statistical classification recognition rates of the SRS are all over 92%. It is clearly seen that the proposed method is capable of the SRS monitoring at various cutting parameters, thanks to its easy feature extraction method, convenience in use and high accuracy. © 2013 American Scientific Publishers All rights reserved.