Fuzzy inference system for modelling machining parameters in electrical discharge machining

Abstract

The monitoring of the machining process has become a preliminary step in the mechanical manufacturing processes, especially for the electrical discharge machining (EDM). However, the monitoring of the electrical discharge machining process remains an ill-defined problem and is generally based on heuristics that are difficult to model. Therefore, artificial intelligence especially the fuzzy logic technique can be applied to processes such as electrical discharge machining in which the experiences and knowledge of the experts play an important role. This study presents a method for the identification of electric discharge machining parameters using the fuzzy inference system. The fuzzy inference system was used to determine the values of material removal rate; surface roughness and radial shrinkage based on process parameter values (open circuit voltage, discharge current, Pulse, duty factor and rinse pressure). The purpose of this is to highlight the development of mathematical models using artificial intelligence (fuzzy logic technique) to correlate relationships of various parameters of the process for determine the influence of those parameters on the quality of the machined parts. The fuzzy model developed for the determination of material removal rate, surface roughness and radial shrinkage formed (fuzzy rules) and tested using experimental data. The mean deviation of the test data did not exceed 3% for the three objectives used, which corresponds to an accuracy of 97%. The results of the tests have shown that the proposed fuzzy model can be used successfully for the selection of the parameters of the electric discharge machining process. The elaborate model offers a more precise and easy selection of EDM parameters.