Intelligent Sensing Unit for Estimation Roughness of Electrical Discharge Machining

Abstract

Estimating the quality of an electrical discharge machining (EDM) workpiece is challenging when attempting to extract features from the stochastic and time-consuming processes. To solve this problem, an intelligent sensing unit for EDM (ISU-EDM) is proposed to extract key machining features for estimating workpiece roughness. During machining, the ISU-EDM simultaneously samples the signals of both the discharge current and voltage, while automatically segmenting the signals according to tool location and discharge effectiveness. Furthermore, the machining features could be extracted from the segmented data by a genetic-algorithm-based distribution fitting method. After applying the features to an automated virtual metrology system, experimental results show that the mean absolute percentage error of roughness estimation is less than 15%.