Machine learning technique ANFIS-COA for enhancing micro-milling performance by investigating the surface roughness and material removal rate

The versatility of micro milling in producing machined components makes it an indispensable instrument in many production sectors, including those dealing with healthcare, optics, automotive, and aircraft. So, measuring the relationship between input and output variables is therefore essential for the efficient operation of the milling process. This study proposes a novel adaptive neuro-fuzzy inference system coupled with a coati optimization algorithm (ANFIS-COA) model to enhance micro-milling performance. The ANFIS-COA model leverages the learning capabilities of neural networks and the optimization power of the colliding bodies algorithm to build an efficient predictive model for surface roughness and material removal rate. The key novelty of the approach is the integration of the COA optimization technique with the ANFIS architecture, which enables robust parameter tuning and optimization of the fuzzy inference system. The objective of this study is to develop a predictive model for estimating surface roughness and material removal rate during micro milling Al6061 alloy based on the Adaptive Neuro-Fuzzy Inference System (ANFIS) and Coati Optimization Algorithm (COA). The name of the hybrid system is ANFIS-COA. The RSM-BBD involves varying process variables such as spindle speed, feed rate, and depth of cut at three distinct levels for evaluating ANFIS-COA. Then, ANFIS-COA is trained using 70% of the exploratory data, while 30% of the data is employed to evaluate its performance. The developed model has been assessed through the evaluation of statistical benchmarks. Moreover, to emphasize the achievement of ANFIS-COA, a comparative evaluation is performed with ANFIS, ANFIS-PSO, ANFIS-GWO, and ANFIS-RSA. According to the results, the ANFIS-COA outperformed the other models by achieving the ideal metrics RMSE (0.0972), MAE (0.0570), MAPE (0.0716), and R² (0.9721) for measuring surface roughness and RMSE (0.0126), MAE (0.0094), MAPE (0.0417), and R² (0.9826) for measuring material removal rate. Eventually, the combination of the ANFIS and COA models could be believed to be a sophisticated system for optimizing milling configuration.