Application of artificial neural networks coupled with an orthogonal design to simulation for powder mixed EDM

Electric discharge machining (EDM) is one of the most accurate manufacturing processes available, and it is quite indispensable to aerospace industries. Powder mixed EDM (PMEDM) is a new process through which a better surface can be reached, nevertheless, PMEDM is a complicated and random process influenced by many parameters, and the intricacies of mechanism hinders its application. This paper presents a method that can be used to forecast the surface roughness in the PMEDM process with the application of artificial neural networks (ANNs). ANNs with back propagation (BP) algorithm coupled with an orthogonal design was applied to the modeling of the PMEDM system and simulation of experimental conditions. An orthogonal design was utilized to design the experimental protocol, in which peak current, pulse-on time, discharging area, servo voltage and no-load voltage were varied simultaneously. The results showed that the ANNs model reflected the complex relationship of PMEDM and had high predicting accuracy