Combining the interacting multiple model method with particle filters for manoeuvring target tracking with a multistatic radar system

Abstract

Practical problems on target localisation and tracking arise in many military and civilian applications. This study investigates a problem on locating and tracking a manoeuvring target for a multistatic radar system. An approach that combines the interacting multiple model (IMM) method with variants of particle filters is implemented. The IMM method accounts for mode switching, whereas the particle filters account for non-linearity and/or non-Gaussianity in the dynamic system models for the posed problem. The approach consists of a constant velocity model, a constant acceleration model and a coordinated turn model. This work considers six combinations of extended Kalman filters (EKFs), unscented Kalman filters (UKFs) and particle filters for the models. A simulation study is carried out to evaluate the implemented IMM algorithm variants. Based on the test results obtained, IMM variants that use computationally efficient particle filters in the coordinated turn models, with EKFs and/or UKFs in the remaining models possess potential to achieve superior accuracy in state estimation for problems with significant non-linearity and/or non-Gaussianity.