Abstract: As an important part of ocean acoustic channel, the geoacoustic parameters of seabed largely determine the characteristics of ocean acoustic transmission. Geoacoustic parameters can be obtained by inversion algorithm, and the range-dependent geoacoustic inversion is a hot research issue in recent years. Particle filtering is an efficient sequential optimization algorithm, and the traditional particle filtering algorithm can be used to estimate the geoacoustic parameters when the seabed characteristics change slowly with range. However, when the seabed characteristics change dramatically with range, such as in the case of stratification change, the traditional particle filtering algorithm may fail. In order to solve this problem, some literatures use the particle filtering with bridging re-sampling for geoacoustic parameters inversion. On this basis, the particle sampling method is further improved in this paper, and the improved method is applied to the inversion of seafloor reflection coefficient. The processing results of the simulation data prove that the improved particle filtering algorithm proposed in this paper still has good performance under the condition that the geoacoustic parameters of seabed change dramatically with range, and can accurately estimate the variation law of seabed characteristics with distance and the geoacoustic parameters of sediment layers.