The core loss estimation of a single phase inverter transformer by using adaptive neuro-fuzzy inference system

In this study, the core losses of a single phase inverter transformer are simulated with finite element analysis (FEA) software according to the certain parameters such as winding turns, primary current and operating frequency values, and then 500 of data are obtained with FEA parametric simulation studies. The adaptive neurofuzzy inference system (ANFIS) model is built by considering the simulation data for estimating the core losses of the inverter transformer. Then, 450 parametric data are utilized for training and the remaining 50 data are used for testing the ANFIS model. The average percentage errors (APE) regarding the estimated core losses for training by ANFIS is obtained as 7.19%. The constructed model is then tested over the test data and the APE value is achieved as 5.04%. The simulation results show that the compatibility of the results obtained with the ANFIS model can be used for the core loss estimation of the inverter transformer.