Tool life prediction is critical to workpiece quality and machining costs. In this paper, a remaining useful life prediction method was proposed, which considering the dimension optimization accuracy of vibration features and the iterative speed of industrial models. Firstly, an optimization method for extracting spatial information of vibration signals was found. The spatial topology of high dimensional vibration information was preserved, when achieving dimension reduction of high dimensional feature. Secondly, a determination method of loss function was proposed to ensure the extraction effects of spatial information. Meanwhile, the k value selection method was proposed which considered to balance the time efficiency factor by using the extraction effect of spatial information. Thirdly, an optimal strategy and model for the remaining useful life prediction of the tool which is suitable for industry was determined, in order to ensure the accuracy and increase the speed. Finally, the superiority and effectiveness of the proposed method were verified by using the processing data of practical tool life cycle.

The traditional predictive model for remaining useful life predictions cannot achieve adaptiveness, which is one of the main problems of said predictions. This paper proposes a LightGBM-based Remaining useful life (RUL) prediction method which considers the process and machining state. Firstly, a multi-information fusion strategy that can effectively reduce the model error and improve the generalization ability of the model is proposed. Secondly, a preprocessing method for improving the time precision and small-time granularity of feature extraction while avoiding dimensional explosion is proposed. Thirdly, an importance coefficient and a custom loss function related to the process and machining state are proposed. Finally, using the processing data of actual tool life cycle, through five evaluation indexes and 25 sets of contrast experiments, the superiority and effectiveness of the proposed method are verified.