Although helically coiled microchannels have a powerful effect on the micro-mixing, they cause a more pressure drop and requires more energy to pump fluids. In this study, Adaptive Neuro-Fuzzy Inference System (ANFIS) was applied to predict the effects of the coiled microchannel geometry on the performance ratio. The mixing performance has been investigated using the Villermaux/Dushman reaction. The performance ratio was calculated by the measured pressure drops to evaluate the mixing and flow characteristics at the same time. Reynolds number, flow ratio and dimensionless parameters including curvature ratio and coil pitch were considered as input data of the model. The results indicated the high accuracy of the model to predict the target data. The mean relative error (MRE) and mean square error (MSE) for ANFIS modeling was 0.71% and 5.72 × 10-4, respectively. The precision of the estimated values for the testing data set (which were not introduced to the model previous training) proved the validity of the ANFIS.
