Abstract: In allusion to the frequent electrical signals distortion, unstable discharge state, and even the discharge state mutation in micro electrical discharge machining (EDM) and deep-hole EDM, the EDM discharge state progressive mapping detection system based on fuzzy logic is established, and local wave decomposition theory is introduced into the study of EDM discharge state prediction. When combined with system identification theory and linear prediction method, the mathematical model of EDM discharge state prediction based on local wave decomposition is established, theory and the corresponding prediction method is put forward. The experiments show the high accuracy, short identification data, fast arithmetic speed, and high real-time performance of this new prediction method. And this method can eliminate the lag occurring in traditional prediction methods. It improves the efficiency of micro and deep-hole EDM, enhances the stability of the prediction and control system, and ensures the process goals, etc.

Key words: Discharge state prediction, Electrical discharge machining, Fuzzy logic, Local wave decomposition, Progressive mapping detection