Ground penetrating radar is widely used in road underground disease detection tasks because of its advantages of convenience, high efficiency and non-destructiveness. However, there is no uniform identification standard for radar images to be recognized, so it is urgent to provide a basis for disease detection. Because GPRSIM ground penetrating radar forward modeling software can be customized the shape of the model, the antenna frequency and the fast speed of forwarding, this study designed disease model of road subgrade void with different filling types, shapes, scales, depths and antenna center frequencies based on GPRSIM. The void disease model have to be carried out forward simulation research, and analyzed the characteristics of radar images. The results have an important guiding role in the interpretation of radar images. At the same time, it is also used to provide an objective theoretical basis for disease identification.