Tracking A Maneuvering Target Using Interacting Multiple Model Algorithm based on Trajectory Function of Time

This paper addresses maneuvering target tracking problem of the follow-up system based on the curve fitting framework. In our work, we extend the trajectory function of time (T-FoT) method to account for maneuvering target tracking field. To tackle the problem that the maneuvering target can not be accurately modeled, we have proposed the interacting multiple model algorithm based on trajectory function of time (IMM based on T-FoT). Our proposal avoids the difficulty for Markov jump (including maneuvering) and process noise modeling, yielding a data-driven tracking approach using little a-priori target model information. Simulation is conducted to demonstrate the effectiveness of IMM based on T-FoT, in comparison with IMM based on KF approach.