Interacting multiple bias model algorithm with application to tracking maneuvering targets

The interacting multiple bias model (IMBM) algorithm is presented as an approach to state estimation for systems with Markovian switching coefficients that can be isolated to a system bias. The IMBM algorithm utilizes the interacting multiple model (IMM) algorithm and recent developments in two-stage state estimation. The IMBM algorithm is well suited for tracking maneuvering targets, where the target acceleration is modeled as a system bias. This algorithm is called the interacting multiple acceleration model (IMAM) algorithm. Simulation results for comparing the performances of the IMM and IMAM algorithms are given, together with a computational count for the two algorithms which indicate that the IMAM algorithm requires approximately 43% of the computations of the IMM algorithm when a constant velocity and two constant accelerations models are used.<<ETX>>