The analysis of dual axis solar tracking system controllers based on Adaptive Neural Fuzzy Inference System (ANFIS) / M.S.I Zulkornain, S.Z. Mohammad Noor, N.H. Abdul Rahman and Suleiman Musa

Artificial intelligence is commonly used in Photovoltaic (PV) control systems. Adaptive Neural Fuzzy Inference System (ANFIS) is one of the intelligent strategies that can be employed in the system controller. ANFIS technique shows high accuracy as it involved several processes which are the Fuzzy layer, Fuzzy Rule layer, Normalization layer, and Output Membership layer. The main objective of the proposed work is to model the dual-axis solar tracker using MATLAB software by utilizing the ANFIS technique, hence improving the performance of the solar system. The data used for training and testing are elevation angle and azimuth angle. 80% of the data is used for training and another 20% for testing in order to predict the solar radiation toward PV panels. A different set of input membership functions (MFs) is used in the system, which are Five MFs, Ten MFs, and Fifteen MFs. These MF are simulated to produce the best prediction of solar radiation. The results show average error gained for both training and testing data and minimum error indicates the accuracy of the predicted angle of dual axis solar tracker. In the finding, overall results show a good correlation between the actual and prediction value with 15 input MFs as it produced the lowest error value.