Multi-objective Optimization in EDM Process Using Backpropagation Neural Network-Genetic Algorithm (BPNN-GA)

Abstract

Surface roughness (SR) and material removal rate (MRR) are performance evaluations of the electrical discharge machine (EDM) Sinking process. They depend on the process variables inputted in the EDM Sinking. In this study, an attempt was made to improve the sinking EDM process to increase MRR and decrease SR in AISI P20 materials with process variables such as electrode type, pulse on time (Ton), pulse off time (Toff), and gap voltage (GV). The Taguchi method’s L18 orthogonal array is employed in the experiment. Analysis of variance (ANOVA) was performed to determine the influence of the process parameters on the response parameters. The experiment was completed with 18 triels with two replicates. The proposed method for modelling and optimization is the Backpropagation Neural Network (BPNN) combined with Genetic Algorithm (GA). The BPNN is developed based on the process variables and the responses measured during the experiments. The developed BPNN model is fed into the GA algorithm. Based on the modelling and optimization results of both methods, an error of less than 5% is obtained, which proves that the hybrid BPNN-GA methods are acceptable.