Neural network modelling and multi objective optimisation of electrical discharge diamond cut-off grinding (EDDCG)

Abstract

This paper presents the development of neural network model and parameter optimisation of electrical discharge diamond cut-off grinding (EDDCG) during machining of cemented carbides for material removal rate (MRR) and average surface roughness (Ra). The experiments were carried out on a self developed electrical discharge diamond cut-off grinding (EDDCG) attachment on conventional EDM machine. The numbers of experiments were selected based on full factorial design of experimental procedure. After experimentation the data set were divided into a training set and testing set for ANN modelling. The ANN back propagation algorithm with four inputs (current, pulse on-time, duty factor, wheel RPM) two outputs (MRR, Ra) and one hidden layer with 15 neurons is proposed to establish the process model. The model after proper training was found to be capable of predicting the response parameters. The predicted results from ANN model are discussed to get the insight of the process. Further, multi-objective optimisation of EDDCG is done using GRA coupled with entropy measurement method.