Tool Remaining Useful Life Prediction Under Variable Milling Conditions Considering Random Failure Threshold

Tool reliability is one of the key factors to ensure the quality and efficiency of machine tool machining. Aiming at the influence of changing milling conditions on the degradation signal during tool machining and the uncertainty of the failure threshold, this paper proposed a method for tool remaining useful life (RUL) prediction under variable milling conditions considering random failure threshold. Firstly, high-dimension features are extracted from multi-sensor signals. The feature that best characterizes tool degradation is selected as a degradation signal through correlation analysis. Secondly, a degradation signal transformation algorithm is designed to eliminate the influence of milling conditions on the amplitude of the degradation signal. Then, the tool degradation model is constructed based on the Wiener process, and the RUL distribution is deduced considering the random failure threshold. The model parameters are obtained using the maximum likelihood estimation (MLE) algorithm to achieve the tool RUL prediction. Finally, the effectiveness of the method is verified by the milling cutter experimental data. The results show that the proposed method can eliminate the influence of milling conditions on the degradation signal and improve the accuracy of tool RUL prediction.