Prediction of flow field in a solar chimney using ANFIS technique

Abstract Solar chimneys have been intensively studied as an effective method for natural ventilation of buildings. Though numerical methods, such Computational Fluid Dynamics (CFD), widely utilized in studies, they usually require extensive computational resources. Moreover, experimental study is quite complicated and costly. In recent years, machine learning has started to be used a tool the thermal-fluid field. this study, order save time cost, Adaptive Neuro-Fuzzy Inference System (ANFIS) technique, class adaptive networks that incorporate both neural fuzzy logic principles, combined with CFD. A simulation model was first validated by experiment from another The result documented dataset using CFD code ANSYS Fluent (Academic version 2020 R2). Then, are train validate ANFIS model. particular, predict fluid flow field 2-dimensional typical solar chimney when heat flux changes range 400 1000 W/m 2 . Inputs position flux, while outputs temperature velocity at location. As result, models could achieve R values 0.997, 0.97 (training set) 0.994, 0.9715 (testing set); RMSE 1.009, 0.00224 1.074, 0.0204 velocity, respectively. Those results acceptable. By model, large amounts fields different scenarios can estimated simultaneously. Therefore, it expected engineers architects quick process design.