ANFIS prediction modeling of surface roughness and cutting force of titanium alloy ground with carbon nanotube grinding wheel

This study harnesses the Adaptive Neuro-Fuzzy Inference System (ANFIS) as a powerful tool for modelling key parameters specifically surface roughness, Metal Removal Rate (MRR), and Cutting forces pertinent to the machining of titanium alloy using a cutting-edge carbon nanotube (CNT)-infused grinding wheel. The investigation is rooted in a meticulously designed experimental framework employing a comprehensive L27 full factorial design, with the crafting of CNT-mixed grinding wheels tailored for the machining process. In the pursuit of accurate predictions, a first-order Sugeno-type fuzzy interference model is strategically employed for the estimation of output parameters. The ANFIS model is subsequently developed, drawing upon machining parameters derived from meticulously curated and trained datasets. Evaluation of prediction performance reveals testing errors that underscore the robustness of the model: 2.13% for surface roughness, 0.15% for MRR, and 4.24% for Tangential force. A comparative analysis of the ANFIS models further highlights their efficacy, showcasing high residual R² values that affirm a robust fit with experimental data in the context of CNT grinding. This underscores the model’s capability to encapsulate the intricate dynamics of the machining process. Importantly, the proposed model exhibits significant potential for real-time estimation of surface roughness in CNT-based grinding applications, suggesting its applicability and utility in advancing the precision and efficiency of machining operations in this technologically advanced domain.