Abstract: With the development of The Times, the application scope of swarm sensing technology in various industries is increasing, and unmanned aerial vehicles (UAVs) with characteristics of low cost and intelligence are also gradually moving toward clustering. Bee colony positioning has become a new solution for scenes with incomplete infrastructure equipment. In order to reduce the complexity of bee colony cooperation, the number of cooperation can be optimized by station selection strategy. However, for moving targets, frequent station selection not only improves the positioning performance, but also leads to the multiplication of computational complexity, that is, there is a contradiction between the positioning performance and the positioning complexity. Aiming at the above problems, this design proposes a target tracking algorithm based on UAV bee colony control (AMP-IMM-EKF). Under the condition of increasing the tracking algorithm module to ensure the positioning performance will not be reduced, it reduces the number of station selection and introduces the interactive multi-model algorithm IMM to make up for the defects of the single model algorithm. At the same time, the transition probability of IMM is updated adaptively in the hive positioning scenario, so as to improve the model matching degree and realize the real motion trajectory tracking. According to the results of comparative experiments, the AMP-IMM-EKF algorithm can shorten the switching time of the model from 10-20 s to about 5 s, which greatly reduces the influence of positioning complexity in the application of the model.