Online modeling and monitoring of power consumption, aerosol emissions, and surface roughness in wire cut electric discharge machining of Ti-6Al-4 V

Estimation of energy consumption in machining plays an irreplaceable role in monitoring and improving energy efficiency with reduced emission and surface roughness. This paper proposed a novel data-based grey online modeling and monitoring system for the power consumption, aerosol emissions, and surface roughness in wire cut electrical discharge machining of Ti-6Al-4 V using the grey theory GM(1, N ). The proposed methodology needs a very less number of data samples for modeling and monitoring with no training time. Amplitude of wire vibration in X - and Y -directions and absolute difference of wire amplitude in data series are inputs for the grey online modeling and monitoring system. The root mean square error for different parameter settings is estimated and identified as an optimum combination of parameters. The combination of amplitude of wire vibration in X -direction and absolute difference of wire amplitude in X - and Y -directions is significant in estimation of power consumption (root mean square error is 0.81) as well as aerosol emissions (root mean square error is 0.75), and the combination of amplitude of wire vibration in Y -direction and absolute difference of wire amplitude in X - and Y -directions is significant in estimation of surface roughness (root mean square error is 0.41). An artificial neural network is also developed and trained with a feedforward backpropagation algorithm that predicted the power consumption, emissions, and surface roughness. A comparison between the grey online modeling and monitoring system and the system developed by the neural network found that grey online modeling and monitoring showed better performance with less training models.