The successful simulation of fuel spray combustion in aero-engines to a certain extent depends on the accuracy of the droplet size distribution used. To achieve high-precision predictions of primary atomization characteristics, this paper proposes establishing a correlation model for the pre-filming airblast nozzle using surrogate model technology. The fluid simulation of the primary atomization of the pre-filming airblast nozzle was carried out by numerical simulation method and compared with the experimental results. The maximum error of the spray angle was only 1.53%, verifying the correctness of the simulation results. Based on the SVR surrogate model method optimized by the improved quantum particle swarm optimization algorithm, a high-precision correlation model between the injection parameters (inflow air pressure, inflow air temperature, fuel temperature, swirler air pressure drop and fuel-to-air ratio) and the primary atomization characteristics (spray angle and Sauter Mean Diameter (SMD)) were established. The cross-influence of injection parameters on the spray angle and the SMD were dis-cussed. The research results show that the SVR correlation model optimized by the improved quantum particle swarm optimization algorithm has higher fitting accuracy and lower prediction error. The fuel-to-air ratio has the most significant influence on the spray angle and the SMD, followed by the inflow air pressure and the swirler pressure drop. Establishing these correlation models not only provide valuable input for subsequent spray combustion simulations but also offer insights into the atomization mechanisms of the pre-filming airblast nozzle under swirling conditions.