Micro-mold insert fabrication in aluminum 7075 using micro-EDMAnthuvan, Naveen (Avtor) Vijayan, Krishnaraj (Avtor) Valentinčič, Joško (Avtor) Mohana Krishnan, M. (Avtor) EDMmicroinsertGRAANNelectrical discharge machininggrey relational analysisneural networkThis work presents the fabrication and optimization of a T-shaped micro-mold insert using Micro-Electrical Discharge Machining. Aluminum 7075 was employed as the work material, while EC – grade copper electrode was used as the tool. Experiments were designed using a Taguchi L9 orthogonal array to vary current, voltage, and duty cycle. ANOVA revealed that current is the most dominant influencing parameter, contributing to more than 80 % of the variation in the output responses, followed by duty cycle, whereas voltage exhibited the least influence. The maximum material removal rate of 36.3696 mm▫$^{3}$▫ /min, minimum electrode wear rate of 0.2837 mm▫$^{3}$▫ /min, and a maximum dimensional accuracy of 92.65 % were achieved. GRA applied to the L9 experimental data identified the optimal parameter combination as 3A current, 50 V voltage, and 8s duty cycle. To improve data reliability, the L9 dataset was augmented to an L27 dataset using the Mix-up data augmentation method. Subsequent GRA and ANN optimization of the L27 dataset consistently confirmed current as the primary influencing factor, followed by duty cycle and voltage. The optimized parameters from the L27 analysis were 3 A current, 40 V voltage, and 8 s duty cycle, which are adopted as the final machining conditions.20262026-07-07 10:13:55Članek v reviji184428sl