OPTIMIZATION OF SURFACE ROUGHNESS AND MATERIAL REMOVAL RATE IN TURNING AISI 420 STAINLESS STEEL USING ANN AND NSGA-II

Abstract

This study investigates the optimization of surface roughness (Ra) and Material Removal Rate (MRR)
in the AISI 420 Stainless Steel turning process using Artificial Neural Networks (ANN) and the Non-
dominated Sorting Genetic Algorithm II (NSGA-II). The ANN model is trained using 22 data points, while
seven are reserved for testing. The ANN model demonstrates excellent predictive performance, with
correlation coefficients (R2
) exceeding 0.95 for both Ra and MRR predictions. The SHapley Additive
exPlanations (SHAP) technique is employed to interpret the ANN model and identify the most influential
input parameters on the predictions. Among the factors affecting Ra, feed rate emerges as the most
significant, followed by depth of cut, cutting speed, and tip radius. Integrating ANN and NSGA-II allows
for effective optimization, resulting in a range of Pareto solutions for Ra (0.414 to 0.942µm) and MRR
(5.546 to 15.577cm3
/min). These findings offer valuable insights for decision-makers in selecting optimal
cutting parameters and demonstrate the potential of ANN and optimization algorithms to enhance
machining processes. Future research can explore broader applications and consider additional factors to
improve optimization accuracy.