Remaining Useful Life Prediction of Cutting Tools Based on Support Vector Regression

Abstract

Abstract Remaining useful life (RUL) prediction of cutting tools is critical to effective condition based maintenance for reducing downtime, ensuring quality and avoiding accidents. In this paper, a RUL prognostic method based on support vector regression (SVR) is proposed for predicting cutting tool’s life. The proposed method consists of two main phases: an off-line phase and an on-line phase. In the first phase, the signal features are extracted from raw data, and then the SVR models with considering different length of signals at past times are established to reflect the relationship between monitoring data and tool life. In the second phase, the constructed models are used to predict cutting tool’s RUL, and the best signal length for accurate prediction result is obtained. The proposed method is applied on experimental data taken from a computer numerical control (CNC) rotor slot machine in a factory. The result shows the validity and practicability of this method.