Classification Model for Real-time Monitoring of Machining Status of Turned Workpieces

Abstract Turning is a molding process widely used in the contemporary machinery manufacturing industry. During the turning process, it is necessary to monitor the machining process of the workpiece in real time in order to improve the surface quality, machining stability and reduce the tool wear cycle. In this paper, a tool chattering state recognition model is designed based on a denoising autoencoders (DAE) feature reduction network and a bidirectional long and short term memory network (BiLSTM). The feature reduction method of DAE is studied, which puts the reduced data into the BiLSTM model for training to reduce the learning difficulty of the network and improves the anti-interference capability. In terms of classification accuracy, the proposed DAE-BiLSTM model provides a high-quality classification of stable processing, transition processing and severe tremors stages in turning chattering state recognition.