Multiple-sensor Fusion Tracking Based on Square-root Cubature Kalman Filtering

Abstract

Nonlinear state estimation and fusion tracking are always hot research topics for information processing. Compared to linear fusion tracking, nonlinear fusion tracking takes many new problems and challenges. Especially, the performances of fusion tracking, based on different nonlinear filters, are obviously different. The conventional nonlinear filters include extended Kalman filter (EKF), unscented Kalman filter (UKF), particle filter (PF) and cubature Kalman filter (CKF), and the recent square-root cubature Kalman filter (SCKF) has been paid more and more attention by researchers because of its advantages of computation complexity and estimation accuracy over other nonlinear filters. However, the SCKF is mainly designed for single sensor system, and for present nonlinear multi-sensor system, it is not applicable. Based on the current results of fusion tracking algorithms, this paper provides a novel multi-sensor fusion tracking algorithm based on the SCKF. Firstly, a brief introduction of basic SCKF algorithm is given. Then, the centralized fusion tracking frame with augmented measurements and the sequential fusion tracking frame are presented respectively. Finally, two computer simulation examples are demonstrated. Simulations results show that the centralized SCKF can obtain better performances of accuracy, stability and convergence.