Application of a hybrid Taguchi-entropy weight-based GRA method to optimize and neural network approach to predict the machining responses in ultrasonic machining of Ti-6Al-4V

The present study was carried out to perform predictive modelling of material removal rate (MRR) and tool wear rate (TWR) during ultrasonic machining (USM) of titanium (Ti) alloy (Ti-6Al-4V) by realizing an optimum artificial neural network (ANN) created by exploring the effect of two different learning algorithms with varied number of neurons in hidden layer. Experimental studies were carried out to explore the effect of various process parameters of ultrasonic machining on response variables MRR and TWR. The basic nature of USM makes these two variables a conflicting one and, therefore, an entropy weight-based grey relational method was used to optimize the process for the two response variables. It was found that the ANN-based predictive results were very closely related to actual experimental findings.