Intelligent Modeling and Optimization of MRR in WEDM of VCN۱۵۰ Using GPR and Genetic Algorithm

Wire Electrical Discharge Machining (WEDM) is an effective non-traditional process for machining hard-to-cut materials with complex geometries. VCN۱۵۰ (۱.۶۵۸۲) steel, due to its high strength, hardness, and wear resistance, is difficult to machine using conventional methods, making WEDM a suitable alternative. This study presents a data-driven modeling and optimization framework to predict and maximize the material removal rate (MRR) during high-speed WEDM of VCN۱۵۰ steel. An experimental dataset was generated using the Box–Behnken design by varying discharge current, pulse-on time, pulse-off time, and wire speed. Gaussian Process Regression (GPR) was employed to model the nonlinear relationship between process parameters and MRR. The model demonstrated satisfactory predictive accuracy with RMSE = ۰.۰۰۲۷۵, MAE = ۰.۰۰۲۰۵, and R² = ۰.۸۴۰۷ using ۵-fold cross-validation. The trained GPR model was coupled with a genetic algorithm to determine optimal machining conditions, resulting in a predicted maximum MRR of ۰.۰۶۵۳ gr/min. Sensitivity analysis identified discharge current and pulse-on time as the most influential parameters.

کلیدواژه ها:

Material Removal Rate (MRR) ، WEDM ، Optimization ، Modeling