Maneuvering target tracking by using particle filter

The aim of this research is to track a maneuvering targets e.g. ship, aircraft, and so on. We use a state space representation to model this situation. Dynamics of the target is represented by system model, firstly in continuous time. Discretized system model is actually used. Position of the target is measured by radar, and this process is described by nonlinear observation model in polar coordinate. To follow abrupt changes of the target's motion due to sudden operation of acceleration pedal, break, and steering, we propose a use of heavy-tailed non-Gaussian distribution for the system noise. Consequently, the model we use here is a nonlinear non-Gaussian state space model. Particle filter is used to estimate the target's state of the nonlinear non-Gaussian model. Usefulness of the method is shown by simulation.