Low Eatht Orbit mega-constellations can provide global seamless coverage of satellite communication services, but also lead to users being covered by multiple satellites. How to choose the optimal satellite for access has become a key issue in the access technology of low-orbit mega-constellations. In response to the above problems, this paper proposes a multi-objective satellite access optimization algorithm based on particle swarm, the objective function comprehensively considers the distance between the satellite and the earth, the remaining service time of the satellite, and the remaining load of the satellite. According to the Quality of Service(QoS) requirements of different traffic types, corresponding weights are assigned to the parameters. Finally, considering the characteristics of a large number of visible satellites and dynamic changes, a low-complexity solution is obtained through the particle swarm algorithm. Simulations show that compared with the comparison algorithm, the proposed algorithm optimizes the trade-off between the distance from the satellite to the earth and the remaining service time of the satellite, guarantees the service time and communication quality, improves the success rate of access, and can flexibly meet the needs of users' different traffic types, adapt to the scenario of users' traffic distribution at different time and spaces, and the scenario of traffic discretization.