A novel particle filtering algorithm based on state fusion
Hang Yu, Yiqun Sun, Qi Ping Cao, G.H. Wang · 2013
To address the limitations of the particle filter algorithm (PF), we propose the fusioned particle filter (FPF). In this new method, the importance density function is generated by state fusion of the extended Kalman filter algorithm (EKF) and the unscented Kalman filter algorithm (UKF). To construct the importance density of samples, the state of each particle is predicted according to the EKF and the UKF, successively. And the feedback of state estimation from the last step is used to implement the update of particles. Thus, using the most of recent measurements and the additional feekback information, FPF can obtain an accurate approximation to the nonlinear non-Gaussian system and as a result, improve the estimation performance. An application example is given to draw a comparison between the FPF and the existing particle filter algorithms. The simulation results show the efficiency of this new approach. (5 pages)