In view of the complex and changeable future battlefield environment， the air defense system needs to realize the assignment of weapons to multiple targets. Because the traditional static weapon target assignment （SWTA） model is restricted by many factors， it cannot adapt to the rapid changes in the battlefield situation. In order to solve the problem of multi-missile dynamic weapon target assignment （DWTA）， the air defense process is discretized into multiple stages， and a mathematical model of DWTA is constructed based on the real-time situation data of the battlefield， and an improved particle swarm optimization algorithm is proposed. Constraints such as the time window for the weapon to fire are considered， and multiple indicators such as interception probability and missile cost are considered in the algorithm. Finally， through a large number of simulation experiments， the rationality and effectiveness of particle swarm algorithm for multi-missile target assignment are verified.