Abstract: Nowadays, with the advancement of propulsion systems, unmanned aerial vehicles(UAVs) have been able to operate in low airspace of urban areas. Due to complex urban environments and risks, there is a pressing need for effective aerial management practices. A UAV trajectory planning method that considers urban risk levels is proposed. A city risk assessment model based on geographic information systems(GIS) was established to analyze safety factors interfering with UAV flights. Considering UAV traversal costs under different risk levels, particle swarm algorithm was improved by introducing contraction factors and traversal coefficients. Simulation experiments were carried out under scenarios of different complexity based on GIS spatial information. The results demonstrate that the algorithm yields superior trajectory quality and the method has generality and versatility.