A Mixture Cauchy Kernel Based Estimation Method for INS/GNSS under GNSS-Challenged Environment
Qingwen Meng, Ming Gao, G. L. Wang, Lu Yang, Hailun Zhang, Jianqiang Wang · 2023
Due to the existence of a challenged environment for intelligent vehicle position such as non-Gaussian measurement noises and sudden changes of system status, the robust forecasting-aided state estimation is pivotal for non-linear inertial navigation system/global navigation satellite system (INS/GNSS) integrated navigation system stability. In order to enhance the robustness of the conventional non-linear method utilized in the INS/GNSS system, this study introduces a robust Cubature Kalman filter (CKF) that relies on the mixture Cauchy kernel (MCK). The utilization of the greatest MCK correntropy is employed as a substitute for the mean square error loss inside a reduced CKF framework. One advantage of employing this approach is the integration of the MCK and CKF, which has been developed in the context of robust information theoretic learning to address non-Gaussian interference in GNSS data, with the CKF, which excels in handling strong model nonlinearities. Additionally, this approach ensures the computational efficiency of the algorithm. Numerical simulation results carried out on INS/GNSS nonlinearity systems validate the efficacy of the proposed methods for state estimation under various types of measurement.