In modern warfare， drone/swarm air raids are gradually replacing the traditional air raid model and building a completely new battlefield situation. Whether the enemy's invasion can be accurately predicted is of decisive significance to the resource allocation， plan planning and efficiency-to-cost ratio control of anti-drone/swarm operations. It is directly related to the resource allocation efficiency and mission success or failure of air defense operations. Based on the above requirements， this paper proposes an air strike scale analysis method for drone/swarm combat： by constructing a multi-dimensional optimization function based on steady-state genetic algorithm， comprehensively considering core indicators such as enemy reconnaissance/strike mission completion rate， reconnaissance coverage rate， combat time and cost， and conducting a global search on the enemy's possible incoming scale marshalling scheme. Through quantitative comparison and analysis of indicators， select the enemy's optimal scale marshalling scheme， and finally form a scientific and effective conclusion on the scale analysis of air strikes.