Supervision of Carbide Tool Condition by Training of Vibration-based Statistical Model using Boosted Trees Ensemble

摘要

Carbide cutting tools form an essential part of the manufacturing industry. A cutting tool, as the name suggests, is a cutting aid, generally harder than the workpiece material which is used for cutting the workpiece material and removing excess material in the form of chips. Any deviation in its condition affects the complete material removal process with respect to quality, accuracy, and durability. Thus, a condition supervision system for fault identification has turned out to be a key priority. The current era of Machine Learning (ML) stimulates the induction of classifier training for tool condition. In this paper, a study on carbide cutting tools is presented during a turning operation carried out on a simple lathe machine. The signature analysis of vibration generated due to the change in the carbide tool condition is carried out. Finally, a Boosted Trees Ensemble is deployed for training of various tool conditions.