Objective Aerosols act as the main carrier of COVID-19 Coronavirus Disease 2019 and the use of random fiber filters can effectively block virus transmission. However due to the complex structure of non-woven random fiber filters it is difficult to establish a three-dimensional model that conforms to the real situation. As a result it is impossible to conduct simulation analysis on them to guide the design and production of the filters. Methods A three-dimensional model of fiber filters with different parameters was established based on random algorithms. Comparative analysis with traditional rule-based models was conducted using various empirical formulas to validate the advantages and effectiveness of this method. The filtration performance of random fiber filters on aerosol particles under specific conditions was analyzed and simulation calculations of pressure loss and aerosol particle capture efficiency were performed using the Euler-Lagrange discrete phase model. Results Simulation experiments showed that fiber filters at fluid velocities of 0. 04 m / s and 0. 112 m / s exhibited stable filtration efficiency for selected aerosol particles ranging from 84. 43% to 95. 28%. Pressure loss in the flow field exhibited a stratified phenomenon with better filtration performance achieved at slower fluid velocities. Conclusion The randomly arranged fiber aggregate model constructed by this method is closer to real fiber materials compared with traditional methods. It provides data and conclusions for understanding and improving filtration technology facilitating the design of better protective measures to prevent virus transmission.