An adaptive interacting multiple model with probabilistic data association filter using variable dimension model

Abstract

A variable dimension probabilistic data association filter is presented to track a single maneuvering target in the cluttered environment. The detection of the maneuver for the model switching is done by the estimation of the acceleration taken from a bias filter of the two stage Kalman filter. An adaptive interacting multiple model probabilistic data association (IMMPDA) filter is also proposed using the variable dimension probabilistic data association filter as one of the target motion model in the conventional IMMPDA filter. A simple Monte Carlo simulation was done to compare the performance of each of the three algorithms.