Prediction Analysis of Surface Roughness Based on Electrical Spark Quality

In this study, the relationship between spark quality and surface roughness of SKD11 die steel in wire electrical discharge machining (WEDM) was investigated. The effects of normal sparks (NS), arc sparks (AS), and short sparks (SS) on the surface roughness of the workpiece were examined. Using a machine learning model optimized with hyperparameter tuning, the surface roughness of the workpiece was predicted by employing three input parameters: NS, AS, and SS counts. The artificial neural network (ANN) model optimized using the tree-structured Parzen estimation (TPE) method was used to predict the surface roughness of the workpiece. The results indicated that the TPE-optimized ANN model exhibited the good performance in terms of mean absolute percentage error (MAPE). The validation MAPE was 1.86%, and the prediction of an additional 10 test groups revealed a MAPE of 1.01%.