Research on discharge state detection of finishing in high-speed wire electrical discharge machine

Abstract

Discharge state detection is a key aspect of the high-speed wire electrical discharge machining (HS-WEDM) control system, as it is strongly correlated with the cutting performance. This paper presents a method for the online detecting discharge state in the process of HS-WEDM based on the voltage and current with the PSO-SVM algorithm. According to the eigenvalues in the regional integral location of the voltage and current waveforms, five discharge states (open circuit, near open circuit, normal spark discharge, near short circuit, and short circuit) that occur in HS-WEDM are discriminated. The proposed classification method has a high accuracy rate and a moderate running time in the identification and classification of discharge state. In test experiments, the above detection system of discharge state was successfully employed to study the finishing characteristics of HS-WEDM in atmospheric and water mist media under different processing parameters. And the spark rate was used as a process index to evaluate the process characteristics of HS-WEDM with surface roughness and cutting speed.