Non Contact Estimation of the Dielectric Properties of Organic Material Using an Inductive RF Sensor and a Multifrequency Approach

摘要

The paper reports on a numerical study aiming at assessing the influence of the frequency on the sensitivity of an inductive sensing method dedicated to the non contact dielectric characterization of organic materials. A three dimensional electromagnetic mesh-free modeling of the interactions between a single loop radio frequency antenna and a dielectric material mimicking a tissue (skin) shows that frequencies allowing a loss factor tanδ close to 1 in the material is particularly relevant to sense its complex permittivity. Moreover, the inverse problem which consists in estimating the complex permittivity of the material starting from the antenna impedance changes is carried out using an artificial neural network. Estimation results shows that a combination of three frequencies providing loss factors in the material such as tanδ = 1, tanδ1 and tanδ<;<;1 enable to significantly enhance the estimation of the dielectric properties, even if sensor positioning uncertainty is involved.