An adaptive variational Bayesian filter for nonlinear multi-sensor systems with unknown noise statistics

• Published in: Signal processing 2021-02, Vol.179, p.107837, Article 107837
• Main Authors: Dong, Xiangxiang, Chisci, Luigi, Cai, Yunze
• Format: Article
• Language: English
• Subjects: Cubature sampling; Information fusion; Inverse Wishart distribution; Uncertain noise statistics; Variational Bayes (VB)
• ISSN: 0165-1684; 1872-7557
• DOI: 10.1016/j.sigpro.2020.107837

•Adaptive filter is proposed for nonlinear systems with unknown noise statistics.•Cubature sampling rule is adopted for nonlinear system state estimation.•Variational Bayesian method adopted to jointly estimate state and noise statistics.•Information filter form developed for multi-sensor distributed feedback fusion.•Simulation and experimental case-studies show effectiveness of the adaptive filter. An improved adaptive variational Bayesian cubature information fusion algorithm for nonlinear multi-sensor systems with uncertain noise statistics is proposed in this paper. Aiming to estimate uncertain process and measurement noise covariances in nonlinear systems, the variational Bayesian theory is combined with the inverse Wishart distribution. System states and uncertain noise covariances are jointly estimated for nonlinear systems by means of cubature sampling, deriving the variational Bayesian cubature Kalman filter (VBCKF-QR). In addition, a variational Bayesian Cubature Information filter (VBCIF-QR) is proposed, and a distributed information feedback fusion algorithm is also derived for multi-sensor systems with unknown noise statistics. Simulation results demonstrate that the proposed VBCKF-QR/VBCIF-QR outperform conventional cubature Kalman/information filters.