At present, the standard of wake separation widely used in China is too conservative, which leads to the limitation of runway capacity. This article takes the approach aircraft wake encounter problem as the research object. On the one hand, it studies the approach wake safety separation reduction method to obtain the reduced critical wake separation. On the other hand, based on the PCA method to reduce the dimensionality of the aircraft performance data in the BADA database, a model hierarchical clustering method suitable for approach wake separation reduction is proposed. According to this method, new model classification results are obtained and constructed Reduced wake safety separation standards. The simulation results show that the model classification method used in this paper has good consistency with the mainstream classification methods at home and abroad, among the mainstream paired aircraft types at Huanghua International Airport, and the approach wake separation standard obtained in this paper can be reduced by more than 0.8km compared with the traditional wake separation standard.