A Variational Bayesian Maximum Correntropy Cubature Kalman Filter with Adaptive Kernel Bandwidth
- Resource Type
- Conference
- Authors
- Li, Zhenwei; Ouyang, Shuaijie; Cheng, Yongmei; Wang, Huibin; Chen, Kezheng
- Source
- 2023 IEEE International Conference on Signal Processing, Communications and Computing (ICSPCC) Signal Processing, Communications and Computing (ICSPCC), 2023 IEEE International Conference on. :1-6 Nov, 2023
- Subject
- Communication, Networking and Broadcast Technologies
Computing and Processing
Signal Processing and Analysis
Adaptive filters
Bandwidth
Robustness
Bayes methods
Noise measurement
Kalman filters
Kernel
nonlinear filter
non-Gaussian noise
variational Bayesian
maximum correntropy criterion
- Language
- ISSN
- 2837-116X
This paper focuses on the estimation problem in the presence of non-Gaussian measurement noise encountered in dynamic systems. The proposed solution is based on a variational Bayesian maximum correntropy cubature Kalman filter with adaptive kernel bandwidth that uses the Mahalanobis distance to adapt in real time the kernel bandwidth. The proposed filter is compared to some recent cubature Kalman filtering approaches, using the univariate nonstationary growth model benchmark. The obtained results demonstrate that the proposed method leads to the estimated values less affected by non-Gaussian measurement noises than other recent cubature Kalman-based filter.