A SELF-LEARNING FUZZY CONTROLLER FOR WIRE RUPTURE PREVENTION IN WEDM

Wire breaking is a serious problem in the application of wire electric al discharge machining (WEDM). A WEDM sparking frequency monitoring an d control system based on the characteristics of the voltage waveform of WEDM is developed. Two types of wire breaking phenomena are categor ised according to the duration of the symptoms: a sudden rise of spark ing frequency, and a sluggish rise of sparking frequency. Based on the se results, a new self-learning fuzzy controller (SLFC) is proposed. T he pulse off-time is regulated in real-time to keep the sparking frequ ency at a safe level so that wire breaking can be prevented. Experimen tal results show that this monitoring and control system can control t he sparking frequency at an optimal high metal removal rate level with out the risk of wire rupture.