Generally, the distance information can't be obtained in maneuver target tracking under the interference environment. In order to get a more accurate tracking result, a new passive tracking algorithm is proposed. Since there is a nonlinear relation between the state variables and the measurements in passive tracking, in the algorithm firstly the least square principle is adopted to pretreatment angle measurements, then the pretreatment result is taken as an input to perform the interacting multiple model filtration. By this, the error brought by the linearization process of the nonlinear measure equation is reduced. However, the accuracy of tracking still cannot be guaranteed because of the localization brought by the fixed structure model muster in interacting multiple model algorithm. In order to enhance the ability to self-adapt to the maneuver target mode, the paper introduces a sequential likelihood ratio test to adjust the model muster, which can reduce the competition among models and ensure the accuracy of tracking. To verify the effectiveness of this algorithm, under the same experiment conditions, using two algorithms to estimate the definite flight track respectively, the simulation shows that the use of the improved algorithm can further improve the accuracy of tracking.