In the paper, facing the high-precision localization demand of formation flight, an Interacting Multiple Model Improved Adaptive Unscented Kalman Filter (IMM-IAUKF) algorithm is proposed for the characteristics of complex maneuverability of aircraft. First, an improved Sage-Husa noise estimator is introduced to estimate the time-varying noise of the measurement in real time, so as to enhance the real-time performance and stability of the filtering algorithm; Then, to address the loss of filtering accuracy caused by the a priori uncertainty of the model transition probability, the model likelihood function is used to adaptively correct the model transfer probability, which improves the model matching accuracy of the IMM algorithm. The experimental results show that the proposed IMM-IAUKF has higher convergence speed and filtering accuracy than the traditional UKF algorithm, which can provide high-precision localization service for formation flight.