Prediction on enhanced electrochemical discharge machining behaviors of zirconia‐silicon nitride using hybrid DNN based spotted hyena optimization

In this work, zirconia composite is machined by the unconventional Electrochemical Discharge Machining (ECDM). Normally, the machining of zirconia is an inconvenient process due to its increased hardness and maximum machining time. The ECDM of zirconia-silicon nitride is done by varying the input parameters such as electrolytic concentration, voltage, and duty cycle. The measured output machined parameters are material removal rate (MRR), Overcut (OC), and Tool wear rate (TWR). In Response Surface Methodology (RSM), the Box Behnken method is used to plan the experimental design of this work. The parameter optimization is conducted using RSM. Besides, the experimented machining performances are validated using hybrid Deep Neural Network-based Spotted Hyena optimization (DNN-SHO) done in MATLAB platform version 2020 a. From the findings, the voltage and electrolytic concentration are identified as signified parameters for improving the ECDM performances from the RSM analysis. The obtained favorable machining performances are 0.371 mg/min of MRR, 162.2 μm of OC, and 0.26 mg/min of TWR. The predicted results from the proposed DNN-SHO for the MRR are 0.402 mg/min, OC is 152.98 μm, and TWR is 0.21 mg/min. The proposed DNN-SHO outcomes are in perfect agreement with the experimented values and are more superior to the RSM, DNN, and DNN-PSO based prediction approaches.