Milling chatter monitoring under variable cutting conditions based on time series features

Abstract

Chatter monitoring is an important task while ensuring machining quality and improving machining efficiency. Under variable cutting conditions, cutting parameters often change, and the cutting state and traditional time-frequency domain cutting features will also change significantly. In order to achieve precise monitoring of chatter under such cutting conditions, this article proposes a time series method named recurrence plot (RP) that can reflect the non-stationary characteristics and state differences of the signal system to analyze the cutting force signal in the cutting process. Firstly, reconstruct the phase space, generate RP; then use recurrence quantitative analysis (RQA) to extract statistical features in the RP that reflect the current state of the system; thirdly, use affinity propagation (AP) clustering method to find out the exemplar features from the RQA; finally, train the exemplar features using the light gradient boosting (LGB) method to obtain the classification prediction model. Experimental results show that this method can effectively identify the machining chatter state and stable cutting state under variable cutting conditions.