A user-friendly fuzzy-based system for the selection of electro discharge machining process parameters

Abstract

This paper introduces a user-friendly intelligent system for the selection of electro discharge machining (EDM) parameters. In this system, a compact selection method based on expert rules, which were obtained from experimental results and extracted from the knowledge of skilled operators, is presented. Expert rules are evaluated by the fuzzy set theory. The developed fuzzy model uses fuzzy-expert rules, triangular membership functions for fuzzification and centroid area method for defuzzification processes. The system was developed on a PC using MATLAB Fuzzy Logic Toolbox. Inevitably, there are many machining parameters (discharge current, pulse duration, pulse interval, gap control, flushing rate, etc.) that should be considered in EDM processes. Selection of these parameters is still an ill-defined problem and generally relies on heuristics, which are not easy to model, and based on the experiences of specialists. In this system, discharge current, pulse duration and pulse interval are the inputs while the outputs are electrode wear, surface roughness and erosion rate. The remaining parameters are considered at constant rate during machining. The system is a compact and homemade tool that can be easily used by an average operator and provides the EDM parameters which lead to less electrode wear, better surface quality and more erosion rate according to the selected operation (finishing, roughing, etc.).