A New Approach to Unequal Dimension States Mixing for IMM Estimator
Nengjie Ou, Shengli Wang · 2019
The interacting multiple model (IMM) estimator has been proven to be of excellent performance and low complexity in tracking agile targets. The success of IMM attributes to mode mixing, where model outputs are mixed for model-conditional reinitialization. Here, a new mode mixing approach is proposed for state estimators with unequal dimensions. The proposed approach adds switching state into target state and uses mode probability and innovation to make the optimal choice online among suggested methods. Compared with the existing methods, the new approach performs better in simulation.