Surface Roughness Prediction and Parameter Selection for Grinding Process with Computer Numerical Control

Abstract

We propose a novel intelligent grinding assistance system (IGAS) for the grinding of silicon carbide (SiC) with computer numerical control (CNC). The proposed IGAS predicts surface roughness (Ra) and suggests suitable parameters for the grinding process. To establish the Ra prediction model, a type-2 functional-link-based fuzzy neural network (T2FLFNN), which updates the network parameter by Lévy-based dynamic group differential evolution (LDGDE), is developed. The LDGDE includes the Lévy flight and dynamic group mechanism to improve the shortcomings of the traditional differential evolution (DE) algorithm. Subsequently, DE is adopted to optimize the grinding parameters according to user requirements. Experimental results of practical machining show that the mean absolute percentage error (MAPE) using the IGAS is as low as 1.62%. Therefore, the proposed IGAS can provide suitable grinding parameters according to the requirements of users.