Maximum Correntropy Extended Kalman Filter for Vehicle State Observation

Qi, Dengliang; Feng, Jianxiang; Ni, Xiaomin; Wang, Lei

doi:10.1007/s12239-023-0031-8

Cited by 7 publications

(2 citation statements)

References 17 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…In the current study of the MCC, there are kernel functions: the Gaussian kernel function and the Cauchy kernel function. The most widely used kernel in correntropy is the Gaussian kernel [26]. On the one hand, the Gaussian kernel function has fewer parameters than the Cauchy kernel function and is easier to optimize.…”

Section: MCCmentioning

confidence: 99%

See 1 more Smart Citation

A MCKF-based cascade vector tracking method designed for ship navigation

Liu,

Ma,

et al. 2024

Meas. Sci. Technol.

View full text Add to dashboard Cite

In ship navigation, the most widely used technology is Global Navigation Satellite Systems (GNSS), which provide the ship’s position and velocity continuously over a period of time. However, when ships are blocked by port buildings or bridges, the quality of signals received from shipborne GNSS receivers may be reduced, resulting in inaccurate ship positioning that poses a risk to navigational safety. In an occluded environment, the measurement process during the signal processing of shipborne GNSS receivers is nonstationary and prone to measurement anomalies, which can contaminate measurement noise with outliers. To address this problem, a cascaded non-coherent vector tracking loop (VTL) is designed, with the Maximum correntropy Kalman filter (MCKF) serving as a cascaded carrier/code pre-filter for shipborne GNSS receivers. The measurement noise covariance matrix of the pre-filter is adaptively calculated and corrected using the carrier-to-noise ratio (CNR) and the maximum correntropy criterion, respectively. The algorithm proposed is more sensitive to outliers than the traditional tracking methods and can effectively solve the state estimation problem under the condition of measurement anomalies. Specifically, the algorithm offers ships with more precise position and velocity estimations and lower signal tracking errors than traditional tracking methods under both static and dynamic conditions, as demonstrated by shipboard experiments. The horizontal positioning is increased by 88.4% and the horizontal velocity error is reduced by 62.1% in the occluded environment under dynamic conditions.

show abstract

Section: MCCmentioning

confidence: 99%

“…Therefore, MCC uses correntropy as a cost function to reduce the effect of heavy-tailed noise. On the assumption that the amount of data available will be limited, the cost function on the basis of the MCC can be further expressed as follows [26]:…”

Section: MCCmentioning

confidence: 99%