Analysis and control of electrical discharge machining (EDM) process

Abstract

A time series of gap states reflecting electrical discharge machining (EDM) process is used to analyze the process. In linear analysis autocorrelation function and power spectral densities are computed to grade machining powers in different sub processes; in nonlinear analysis surrogate data method is used to prove the deterministic nonlinearity of a sub process, an efficient machining process. The deterministic nonlinearity of the process educes the possibility of building a timely varied linear model, approximating the varied gap states. Based on real-time estimated parameters of the model, by using minimum-variance control strategy, a self-tuning regulator is designed to automatically regulate electrode down-time so that gap states follow a specified gap state with the purpose of much faster and more stable machining than an open-loop machining without an adaptive control.