ANN-NSGA-II dual approach for modelling and optimization of process parameters for performance improvement in vertical turret milling machining

Machining parameters optimization is most important factor to get enhanced productivity, quality and sustainability in manufacturing processes. Conventional single-objective approaches of optimization often fail to address the complex multi objective nature problems. In this study, multi objective optimization framework is implemented that includes Grey Relational Analysis (GRA), multiple linear regression, artificial neural networks (ANN), and Non-dominated Sorting Genetic Algorithm II (NSGA-II). These methods were used to optimize vertical turret milling machine performance. The output parameters such as material removal rate (MRR) and surface roughness (Ra) were considered as key responses parameters. The methodology begins with GRA for initial multi-objective parameter screening using a full factorial design with 81 experimental runs on EN24 steel. The regression analysis and ANN modeling were used to capture both linear and nonlinear relationships between input parameters such as spindle speed, feed rate, and depth of cut and output responses are material removal rate and surface roughness. NSGA-II method implemented to optimization for global parameter selection. The results demonstrate that GRA successfully identified optimal parameters with verification experiments showing surface roughness while the ANN model achieved high prediction accuracy for MRR (>90% in most cases) and variable performance for Ra, and NSGA-II provided Pareto-optimal solutions balancing multiple objectives with validation experiments confirming less than 5% error from predicted values.