USING ARTIFICIAL NEURAL NETWORKS TO IMPROVE THE EFFICIENCY OF TRANSFORMERS USED IN WIRELESS POWER TRANSMISSION SYSTEMS FOR DIFFERENT COIL POSITIONS

Abstract

This study uses magnetic resonance-based coupling theory to study the various placements of transmitter and receiver coils in wireless power transfer (WPT) systems. Various coil placements are examined to show where high efficiency can be achieved within the air gap. Basic characteristics such as self-inductance, mutual inductance, and coupling coefficient were calculated. Artificial neural networks (ANNs) in WPT are a powerful technique for predicting performance characteristics. Using ANNs provides an excellent method for streamlining the design process and reducing time-consuming calculations. To quickly determine and optimize coil design, this study compares recent research on ANN applications in WPT and the performance of different types of ANNs in WPT systems. An artificial neural network (ANN) was trained to predict the magnetic properties of a wireless power transfer (WPT) device. Appropriate cost functions have been implemented to train the ANN properly. It was shown that the trained ANN can effectively reproduce the data obtained by the finite element method (FEM). The results show an effective power transmission at different coil placements, with decreased efficiency observed after a certain distance. These data will help determine the proposed WPT system's air gap and angular limits.