Cauchy Kernel-based Maximum Correntropy Extended Kalman Filter for Cooperative Localization of Multi-UUVs
R.J. Li, Hongli Xu, Rong Zheng · 2024
Cooperative localization (CL) of multiple unmanned underwater vehicles (Multi-UUVs) has recently been a popular research topic. Measurement outlier, which impacts state estimation and control of multi-UUVs, is a serious obstacle for CL of Multi-UUVs. Measurement outlier is aberrant data that depart from the typical range when the sensor is sampling data. For solving this challenge, this study develops a Cauchy kernel-based maximum correntropy extended Kalman filter. Firstly, the CL system of Multi-UUVs is modelled. The CL system model is made up of the motion model and the measurement model. The motion model adopts 2D kinematics model, and the measurement model is based on relative distance and relative orientation. Secondly, converting the CL system model to a linear regression problem. Iterative least square method is selected to solve the problem with a loss function based on the maximum correntropy with Cauchy kernel, this robust loss function effectively mitigates the influence of measurement outlier, enhancing the overall performance of the CL algorithm. Finally, by setting the occurrence of continuous outlier and random outlier events, the superiority as well as effectiveness of the proposed algorithm are analyzed and verified by comparing with the contrast algorithms. This study has positive implications for the application of CL of Multi-UUVs.