Study on the Prediction of Surface Roughness of ZTA Ceramics under Ultrasonic Assisted Grinding

Abstract

By means of Adaptive Neural-Fuzzy Inference Systems(ANFIS), the prediction model of surface roughness of zirconia toughened alumina ceramics (ZTA ceramics) under ultrasonic assisted grinding is established, and the model may obtain the higher forecast precision (81.25%) by dealing with nonlinear problem among grinding parameters, and fuzzy rule curved face formed by prediction model showed four input factors have different influence on surface roughness Ra, and they are abrasive grain size, grinding method, grinding depth, axial feed and speed of worktable from top to bottom. So the prediction model provides a new and efficient method for intelligent information processing, and it could be applied in modeling and real-time control etc, and it is possible to be widely applied in engineering.