Artificial Neural Network Based Load Estimation in Single-Input Single-Output Inductive Power Transfer Systems

In inductive power transfer (IPT) systems, load resistance is an important information that is required to perform parameter adjustment at the transmitter side for performance optimization. In this paper, an artificial neural network (ANN) based method is proposed to accurately estimate the load resistance in single-input single-out IPT systems under the presence of variations in mutual inductance between the transmitting coil and the receiving coil. Data to train and evaluate the proposed ANN model are extracted from LTSpice-based simulations. Furthermore, the IPT system is designed to operate in two modes, including power transfer mode and estimation mode. The IPT system is switched from power transfer mode to estimation mode to collect data for estimating load. Experiments are carried out to confirm the effectiveness of the proposed load estimation method. Experimental results show that although coupling coefficient alters from 0.1 to 0.2, the proposed ANN model can estimate loads with mean absolute error of 0.60, mean square error of 0.61 and coefficient of determination of 0.98.