Intelligent diagnostic technique of machining state for grinding

Abstract

Successful grinding of a final product depends upon a large number of parameters that affect the grinding result and are strongly interlinked. It is, therefore, difficult to detect directly the generation of grinding faults such as chatter vibration and burning. In this paper, to achieve the development of an intelligent diagnostic technique for chatter vibration and burning phenomena on grinding process, acoustic emission signals were processed and signal parameters of the acoustic emission were also determined. In addition, a neural network was used as a diagnostic technique of the grinding state. A momentum coefficient, learning rate, and structure of the hidden layer were determined during the iterative learning process and the performance of the diagnostic technique was evaluated.

A Model-Based Computationally Efficient Method for On-Line Detection of Chatter in Milling

“…Prior work on chatter detection generally employs three types of signal processing methods, including: (1) transform domain analysis such as the Fourier transform, power spectrum, the short time Fourier transform (STFT) [7][8][9][10][11][12][13][14][15][16][17][18][19], and wavelet transform [20][21][22][23][24][25][26][27], (2) time domain modeling and analysis [28^6], and (3) pattern recognition [47] and classification algorithms such as artificial neutral networks [25, [48][49][50][51][52][53][54], fuzzy logic [55], the hidden Markov model [56,57], support vector machine [22,57] and the index based reasoner [58].…”