Investigation of Composite Metal Discrimination Using Coil Impedance Characteristics with Machine Learning

Abstract

Metal discrimination is a technology in increasing demand for recycling and other purposes that can be performed by using different methods. The eddy current method utilizes changes in the impedance characteristics of the sensor coils due to eddy current in target. The amount of change on the conductivity and magnetic permeability of the target that can be used for metal discrimination. In this study, we combined the eddy current method with machine learning to discriminate composite metal . In experiments, we successfully discriminated between cupronickel plates and a cupronickel–nickel–cupronickel composite material. Our approach can potentially be applied in sensors to discriminate metals within metals.