Chatter prediction in turning process of conical workpieces by using case-based resoning (CBR) method and taguchi design of experiment

The detection of undesirable vibrations in turning operations is an important task for a manufacturing engineer. Many different approaches were used to chatter detection on turning processes though they need complex analysis and expensive equipment. Case-based reasoning (CBR) is a method that seems to solve many problems addressed in knowledge-based (KB) systems. CBR includes nonlinear calculating units called nodes. The data used for the system performance derived from experiments conducted on a turning machine according to the principles of Taguchi design of experiments (DOE) method. In this research, image processing and machine vision (IPMV) approach and CBR have been carried out to chatter detection. In CBR, by using a large number of nodes on the systems, 4 M Omega quartz crystal oscillator and increasing sampling rate from 7.5 to 30 K/s accurate system could be obtained with 6.55% mean square error (MSE) in comparison to IPMV and neural network that has 14.60% and 8.5% MSE, respectively. Also another special ability of the CBR system is detecting chatter from two materials simultaneously by separate input signals into two materials and downsampling the output of both filters through those materials by using low-pass filter.