Reliability estimation for cutting tool based on logistic regression model

As an important part of CNC machine, cutting tool reliability plays an important role to the total manufacturing effectiveness and stability of equipment. With an accurate identification of cutting tool wear state, a reliability estimation method based on logistic regression model is proposed to reduce tools costs and guarantee a certain surface machining quality. During the manufacturing process, tool vibration signals are measured on-line by experiment. Wavelet packet (WP) transform is employed to decompose vibration signals in order to find out the feature frequency band. Correlation analysis is used to extract the salient feature parameters, which are composed of energy, energy entropy, and time-domain feature. Combined with tool state, a reliability estimation model based on logistic regression is set up and applied to evaluate the reliability indices of the other in-used tool. Under different failure threshold, the reliability and failure time are all estimated accurately. In this method, the operation condition information of equipment is introduced into reliability analysis to reflect the asset time-varying characteristics. It is not necessary to make much assumption about degradation path and distribution function of condition feature.