Adaptive neuro-fuzzy inference system (ANFIS): modelling, analysis, and optimisation of process parameters in the micro-EDM process

ABSTRACT

Micro-Electro Discharge Machining (micro-EDM) is used to machine micro-holes on silver plate of 350 µm thickness using a silver tool of 450 µm diameter  by varying three influencing input process parameters - voltage (V), Capacitance (C), and pulse on-time (Ton). The output responses of interest are Material Removal Rate (MRR), Tool Wear Rate (TWR) and Diametral OverCut (DOC). It has been noticed that the volume of material removed from the electrodes decreases with an increase in depth, which follows the nonlinear behavior. Mathematical modeling is hence, a difficult task. To overcome this difficulty, the simulation model using the Adaptive Neuro-Fuzzy Inference System (ANFIS) with Principal Component Analysis (PCA) has been developed and analyzed. The process parameters are considered as input to the architecture and output response is generated. Sugeno fuzzy model is used to generate fuzzy rules for a given set of data. The predicted values for MRR, TWR and DOC are found to be in the error percentage of 8.67, 3.20 and 13.44 respectively. The quality of machined holes is analyzed using optical microscope.

KEYWORDS:

• Micro-EDM
• material removal rate
• tool wear rate
• ANFIS
• PCA

Disclosure statement

No potential conflict of interest was reported by the authors.