Abstract: Because of its nondestructive and high efficiency, GPR has gradually become an effective detection method in the fields of road detection and underground exploration. However, the traditional GPR thickness inversion algorithm cannot effectively detect parameters such as thickness and relative dielectric constant of layered media, and has a large inversion error of multi-layer media parameters. Therefore, an optimization algorithm based on layered media parameters inversion algorithm is proposed in this paper. From the perspective of time domain, one optimization model based on genetic algorithm were designed with the common middle point method to retrieve the parameter information of underground layered media structure. Through experimental simulation, the method proposed in this paper can be applied to underground multilayered medium, and accurately invert the parameter information of layered medium.