Aiming at the problem that the target motion state changes the accuracy or even the divergence of the target tracking algorithm

an object tracking algorithm based on interactive multiple model（IMM） is proposed.First

the IMM improvement algorithm combines the non-maneuver model with the Singer model.The Singer model can better adapt to the strong maneuvering target by changing the maneuver frequency and acceleration variance.Secondly

for the problem of low accuracy of initial positioning data

the time difference of arrival（TDOA）algorithm is used to estimate the positioning data.The TDOA algorithm has higher precision and lower computational complexity in the line-of-sight environment.The experimental results show that compared with the traditional IMM filtering algorithm

the improved IMM algorithm can effectively improve the positioning accuracy and achieve stable tracking of the target.