Enhanced integral equation modelling for moisture sensors

Abstract

Integral equation (IE) methods are viable for modelling the interaction between microwave signals and porous materials for measurement of moisture content. Practitioners of IE methods frequently use delta basis functions to represent the field in each cell, but the inherent approximations limit accuracy. Although higher order basis functions provide one alternative, here we retain the geometric and cell density advantages delta functions, but address accuracy by correcting for the field integration errors of the conventional moment method. Our approach circumvents the errors that arise when assuming that the field of a cell is adequately represented by a source at its centre, and has resulted in close agreement between empirical results and the model. The new technique retains the rapid calculation and suitability for use in dielectric tomography and for modelling the spatial response of moisture sensors.