In this study, we focus on exploring the propagation characteristics of particle swarms in social networks and analyze the diffusion process of viruses among populations based on system dynamics. The article mainly discusses three propagation influence mechanisms, including individual attributes, group attributes, and particle swarm attributes, and delves into the modeling of diffusion processes based on network structures. Firstly, we adopt the main roads in the transportation network (hub nodes) as the initial network backbone. On this basis, by introducing branch networks with small-world characteristics and scale-free characteristics, we construct a transportation network that integrates multiple properties. Using this network, we conducted a detailed simulation and analysis of the COVID-19 transmission process and compared and verified it with the infection dynamic data of COVID-19 in Shanghai from March to September 2022. The verification results reveal that our proposed model can significantly improve prediction accuracy. Compared with other existing dynamic models, our model demonstrates excellent performance, possessing high practical application value. This study provides robust theoretical support for the propagation characteristics of particle swarms in social networks and lays the foundation for further research and application in related fields.