New Perspectives in Artificial Intelligence-Based Object Detection for Wireless Power Transfer Systems

This paper proposes a possible approach for object detection in Wireless Power Transfer (WPT) systems using artificial intelligence techniques. The main objective is to identify the presence of foreign objects between the primary and secondary windings by processing voltage and current measurements. To achieve this goal, different machine learning methods are proposed and compared. One of the essential features of the proposed method is to avoid the use of additional differential voltage coils, whose effectiveness is affected by the misalignment between primary and secondary coils. Therefore, this work proposes a theoretical approach for identifying foreign metallic objects without additional sensors, sensing coils or expensive equipment.