Milling tool wear prediction using spindle motor current signal

摘要

Machine tool automation requires monitoring techniques to ensure the condition of the machine and the quality of the workpiece. In the present work, an artificial neural network (ANN) and a least-squares support vector machine (LS-SVM) are used to predict the flank wear (VB) of a millingtool (KC710) in the milling operation. The root-mean-square (RMS) value of the spindle motor current is the input for the model, whereas flank wear is taken as the output from the model. The influence of different operating conditions (such as the feed rate, the speed and the depth of cut)on the flank wear of a milling tool has been investigated. It is observed that the LS-SVM provides the smallest value of average error in tool wear prediction among all of the studied approaches. This study shows that the LS-SVM is a robust model for real-time tool wear prediction. Continuousmeasurement of the motor current signal is a relatively easy task and the consequential accurate prediction of tool wear using the motor current signal will help in the automation of the milling process.