Due to the fact that currently the commonly used deep-sea plankton sampling method is to use water pumps for extraction of the seawater to catch plankton through filtration, taking plankton filters as the research object, a three-dimensional fiber membrane microscopic model has thus been established through random algorithms, followed by a numerical simulation of the fiber membrane model based on similarity criteria. Meanwhile, the Lagrangian discrete phase model (DPM) is used to track the motion trajectory of plankton particles inside the filters. The research results indicate that changes in inlet velocity have a significant impact on the pressure and velocity distribution of the flow field. As the inlet velocity increases, the velocity difference inside the flow field increases, with a clear stratification in the pressure distribution in the flow field. Moreover, the pressure loss increases with the increase of inlet velocity. The fiber model with smaller pores is characterized with a better filtration performance, with the filtration efficiency of plankton particles being basically stable at 80%.