<i>q</i>-R<i>é</i>nyi Kernel Functioned Kalman Filter for Land Vehicle Navigation
Jiangnan Xing, Tao Jiang, Yingsong Li
Abstract
In this brief, a q-Rényi kernel functioned Kalman filter (qRKFKF) is proposed based on the q-Rényi kernel function which is to provide better flexibility and performance in non-Gaussian environment. The concrete realization of the proposed qRKFKF are created and analyzed in detail, and its performance is presented and discussed. Three examples are carried out to verify the performance of the proposed qRKFKF for land vehicle navigation via numerical simulations: large outliers in non-Gaussian noise, alpha-stable noise and diverse noise distributions that combines Gaussian and Laplace noises. Compared with Kalman filter (KF), Maximum Correntropy Kalman Filter (MCKF) and minimum error entropy KF (MEE-KF), the qRKFKF is superior to others for combating non-Gaussian noises.