High surface roughness hinders the flow of fluids in microchannels leading to low accuracy and poor-quality products. In this work, the adaptive neuro-fuzzy inference system (ANFIS) was used to examine surface roughness in CO2 laser fabrication of microchannels on polymethyl methacrylate (PMMA). The PMMA substrates were coated with a 500 nm layer of 99.95% pure aluminium. The inputs were speed (10, 15, and 20 mm/s), power (1.5, 3.0, and 4.5 W), and pulse rate (800, 900, and 1000 pules per inch) while the output was surface roughness. A 3-level full factorial design of experiments was used, and 27 experiments were conducted. Using the gaussian membership function (gaussmf), the ANFIS model was developed using the ANFIS toolbox in MATLAB R2022a. Analysis of variance was performed to examine the significance of the inputs. Power is the most significant followed by speed and pulse rate. The mean relative error (MRE), mean absolute error (MAE), and the correlation coefficient (R) were used to examine the accuracy and viability of the model. MRE, MAE, and R were found to be 0.257, 0.899, and 0.9957 (R2 = 0.9914) respectively. The root mean square error (RMSE) was 0.0022 and 3.6099 for the training data and checking data respectively. Hence, the developed model can predict the values of the average surface roughness with high accuracy.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.