Design and comparison of mode-set adaptive IMM algorithms for maneuvering target tracking

The variable-structure multiple model (MM) state estimation approach is utilized for maneuvering aircraft (MA) radar tracking. Two adaptive time-varying mode-set interacting multiple model (IMM) tracking filters are designed and investigated: switching grid (SG) IMM and adaptive grid (AG) IMM algorithms, By Monte Carlo simulations the performances of the algorithms and the respective fixed grid (FG) IMM filter are evaluated and compared over different flight scenarios. It is shown that for the considered specific maneuvering target tracking problem, the SGIMM and AGIMM tracking filters significantly outperform the corresponding fixed structure version (the FGIMM filter) with respect to performance-to-computational load ratio.