Modelling and prediction of surface roughness in end milling operation using adaptive neuro-fuzzy inference system

Abstract

Surface roughness is a widely used index of product quality and the quality of the surface plays a very important role in the performance of the milling operation, as a good quality milled surface significantly improves fatigue strength and wear resistance. Therefore, modelling and prediction of surface roughness of a workpiece in milling operation plays an important role in manufacturing industry. This paper proposes an adaptive neuro-fuzzy inference system (ANFIS) for modelling and predicting the surface roughness in end milling. Three cutting parameters, i.e., spindle speed, feed rate and depth of cut, those have a major impact on surface roughness were analysed. Three different membership functions namely, triangular, trapezoidal and bell-shaped, were used during the hybrid-training process of ANFIS in order to compare the prediction accuracy of surface roughness by the two membership functions. The predicted surface roughness values obtained from ANFIS were compared with experimental data and multiple regression analysis. The comparison indicates that the adoption of above three membership functions in ANFIS achieved much better accuracy than multiple regression model. The bell-shaped membership function in ANFIS achieves slightly higher prediction accuracy than other membership functions.