Research On Detection System Of The WEDM Process Discharge Status Based On Neural Network

Wire cut electrical discharge machining(WEDM)technology has been widely used in processing the difficult-to-machine material,parts required complex contours,special tools and mould manufacturing etc.,and which has played an important role in many areas of industrial production as aerospace,war industry and mould manufacturing.The real-time detecting the discharge status of electrical discharge machining can provide a reliable basis for control of servo feed and optimization of processing parameters,so as to maintain a good status of gap discharge,to achieve efficient and stable processing.Therefore,the discharge status of WEDM can be used as another important indicator to optimize machining parameters.There is a great significance for improving the machining accuracy and cutting speed to research on the high performance detection stability system of WEDM.In this paper,on the base of consulting a large of literatures,from the traditional threshold methods to artificial intelligence-based detection methods to detect the state of discharge were comprehensively analyzed.Due to the randomness and complexity of EDM discharge mechanism,the discharge status of real-time detection technology based on neural network is proposed to identify WEDM discharge status.Usually the discharge status includes open status,normal discharge and short status.In the analysis of WEDM discharge status,the partial open status and partial short status are proposed.A longer period of partial open pulses or partial short pulses may cause that the following pulses become open or short pulses,which will affect the processing.So detecting the partial open and partial short discharge status in WEDM will help to achieve the best control of electrical discharge machining.Clearly based on the neural network theory,using of Hall sensors and high-frequency data acquisition card to complete the WEDM gap signal acquisition has been studied;digital signal logic processing has been built in LabVIEW virtual platform and combined with selected hardware to build the stability detection system,which can effectively detect five kinds of discharge condition:open,partial open,normal spark,partial short and short.This paper has done the experiment about the effects of electrical parameters on discharge status such as pulse width,pulse interval ratio and short-circuit peak current.The waveform using oscilloscope synchronous sampling will be contrasted with signal of detection system to verify detection system reliability.Finally,the application of the detection system in fluid roughing and liquid finishing,gas finishing and water mist finishing,it has been obtained the stability features of different media and spark rate distribution.