An Interacting Multiple Model Tracking Algorithm Based on Rigid-Body Model of Target

Rigid-Body Model (RBM) is a more precise method for motion description of airborne maneuvering target compared with traditional Particle Point Model (PPM). It includes a pose-angle vector in the target state variable vector to model target motion more exactly, which has the potential to improve the tracking performance. Traditional RBMs are usually too complex to be accomplished in practical application. Therefore, a simplified application-oriented RBM is proposed in this paper by analyzing the motion characters of maneuvering targets. It is then used in the Interacting Multiple Model (IMM) algorithm, where the fuzzy association measure between the pose-angle measurement of a target and the current motion mode is calculated, and the association result is used to fuse with the model probability. Numerical simulation results show that the proposed method could effectively improve the tracking performance of IMM because of the enhancement of identification ability of motion model, and nearly bring no additional computational burden.