Inverse Analysis for Inhomogeneous Dielectric Coefficient of Pavement Material Based on Genetic Algorithm

The key of ground penetrating radars application lies in the calculation of dielectric coefficient. The pavement materials are inhomogeneous medium in fact, the particle surface can induce the scatter and diffraction of electromagnetic wave. The inhomogeneous dielectricity can change the characteristics of reflected wave. It may even cause background noise of reflected signal, which will lead to mistakes in signal interpretation. Therefore it is necessary to analyze the inhomogeneous dielectric coefficients by GPR. This paper proposes the solutions of inverse analysis for inhomogeneous dielectric coefficients of pavement materials used GPR data. Two examples are given to assess the validity of genetic algorithms in inversion of pavement materials inhomogeneous dielectricity. The results show that genetic algorithm can converge into true solutions well. The backcalculated inhomogeneous dielectric coefficients can help to evaluate pavement properties further.