THERMAL ERROR MODELING FOR REAL-TIME ERROR COMPENSATION

A modelling strategy for the prediction of both the scalar and the pos ition-dependent thermal error components is presented. Two types of em pirical modelling method based on the multiple regression analysis (MR A) and the artificial neural network (ANN) have been proposed for the real-time prediction of thermal errors with multiple temperature measu rements. Both approaches have a systematic and computerised algorithm to search automatically for the nonlinear and interaction terms betwee n different temperature variables. The experimental results on a machi ning centre shore, that both the MRA and the ANN can accurately predic t the time-variant thermal error components under different spindle sp eeds and temperature fields. The accuracy of a horizontal machining ce ntre can be improved through experiment by a factor of ten and the err ors of a cut aluminium workpiece owing to thermal distortion have been reduced from 92.4 mu m to 7.2 mu m in the lateral direction. The dept h difference due to the spindle thermal growth has been reduced from 1 96 mu m to 8 mu m.