A Reinforced k-Nearest Neighbors Method with Application to Chatter Identification in High-Speed Milling

Chatter is a kind of self-excited vibration which will destroy the manufacturing process badly. The detection or identification of chatter is attracting considerable interest for several years. In this article, a chatter identification method called reinforced k-nearest neighbors is proposed to realize both chatter identification and model self-learning. We conducted large amounts of experiments on a computer numerical control milling machine with different types of sensors in high-speed milling processes, where chatter occurs frequently. Signals from different sensors are compared and features are extracted by statistical methods. Then, a dimensional reduction method t-distributed stochastic neighbor embedding is used for extracting sensitive information and visualization. Finally, the proposed reinforced k-nearest neighbors is used for chatter identification under different cutting conditions and the experiment results show the effectiveness of the proposed method.