Friction compensation of a retrofit milling machine through a fuzzy logic controller combined with neural network-based feedforward control

Abstract

Conventional milling machine contained nonlinear time-varying dynamics because of the effect of irregular friction and obvious backlash inherited in the ball-screw driven table. In this paper, a fuzzy logic controller combined with neural network-based (NN) feedforward control was proposed to improve the contouring accuracy of this milling machine through friction compensation. Experiments on contouring control were conducted for various paths under milling force. The experimental results showed that this control scheme improved not only the contouring accuracy significantly, but also the disturbance rejection and robustness, comparing to the conventional control scheme.