Study On Detection Of WEDM Gap Discharge Status Based On PNN Optimization

WEDM is an excellent non-traditional processing method,it's advantages include cutting workpieces with complex shapes or hard-to-cut materials through the medium,and ensuring better processing quality.In recent years through,the application of water mist,the wire cutting have made great progress in the field of the multiple cutting.It's cutting quality is better than the single emulsion medium and the water spray medium will not cause environmental pollution.Therefore,the cutting properties of water spray medium have research value.The change trend of the corresponding processing result parameters is obtained by detecting the WEDM gap discharge status.Then based on this analysis,the corresponding cutting characteristics and rules are analyzed,which plays a further guiding role in optimizing the processing parameters to improve the processing efficiency,processing quality and processing mechanism.In this paper,the waveform of the gap electrical signal is obtained by the actual cutting experiment in water spray medium.Based on the selected representative waveforms,the voltages and current waveforms in different states during in the machining process are analyzed.The typical discharge state waveforms and according to the waveform are divided into five classes,the discrimination criteria of wire cutting discharge state are established according to the waveform.Through the comparison and analysis of the processing mechanism of cutting in the emulsion,gas and water spray medium,putting forward the thresholds setting of different medium to ensure the accuracy of the testing.According to the classification characteristics of the state,PNN neural network with good phase classification is used as the basis of the detection system program,and the qualified algorithm is obtained by program learning and training combined with the voltage and current signal values of the experimental data of the previous study.The design of the detection system covers software and hardware aspects.Based on experimental conditions and parameters,the appropriate acquisition elements are selected to realize the collection and transmission of electrical signals.The real-time detection of the discharge state of the gap is realized by the functions and programs of the LABVIEW virtual platform design driver software.Incorporating the process prediction module can enhance the system functionality,and a better main interface of the detection system is designed.Finally,the detection system is used to test the results of multiple cutting experiments with different parameters.Two kinds of finishing experiments are carried out.One is the cutting test of different medium with the same thickness,the experimental results are used to study the cutting characteristics of different medium similarity and difference.The other is using 15 ml / min steam spray medium to cut thickness workpieces with single factor experiments to study the influence of five major processing parameters(pulse width,peak current,pulse interval ratio,offset and table no-load speed)of finishing on the three processing quality assessment indicators(surface roughness,cutting speed and spark rate distribution).It can guide the improvement of process indicators.