Abstract:

A class of expected-mode augmentation (EMA) algorithm based on likely-model set (LMS) is proposed. The effective model set is divided into LMS and expectedmode set at each time step. The LMS is decided according to the posterior model probabilities and the topology, and the expectedmode set is generated by the weighted combination of the likely-model set. Then, the target state is estimated based on the union set of the two model sets. The algorithm maintains the precision of the EMA algorithm and reduces the computational load. Simulation results validate the performance of the algorithm.