Multirate interacting multiple model particle filter for terrain-based ground target tracking

Ground target tracking is a nonlinear filtering problem when it incorporates terrain and road constraints into system modelling and uses polar coordinate sensing. Furthermore, when tracking ground manoeuvring targets with an interacting multiple model approach, a non-Gaussian problem exists because of an inherent mixing operation. A multirate interacting multiple model particle filter (MRIMM-PF) is presented to effectively solve the problem of nonlinear and non-Gaussian tracking, with an emphasis on computational savings. The sample subset of each mode is updated at a different rate and mode switches are performed according to a Markov chain at a low rate. For a fixed number of samples, simulation results show that the MRIMM-PF significantly reduces computational costs, with comparable tracking performance to multiple model particle filter.