Seamless integration of polarization compass and inertial navigation data with a self-learning multi-rate residual correction algorithm

Abstract

Yaw angle measurement by inertial navigation systems (INS) is of fundamental importance for military and civilian applications of unmanned ground vehicles. To reduce the INS accumulated error, INS and other navigation methods are commonly combined through information fusion schemes such as Kalman filters. As well, yaw angle measurement can be realized by bionic-vision navigation tools, especially the polarization compass (PC). However, the realization of PC/INS integrated navigation systems is restricted by differences of the PC and INS sampling frequencies, as well as the PC environmental disturbances. In this paper, a cubature Kalman filter with a multi-rate residual correction algorithm is proposed to fuse INS and PC navigation outputs with different sampling frequencies. As well, a long short-term memory network is employed for PC output prediction and fusion when actual PC data is not available. The performance of the proposed method has been experimentally verified using PC and INS data for different trajectories of ground-vehicle navigation tests. Comparative results of the proposed and traditional methods show a superior accuracy of the proposed method for the yaw angle measurement fusion even when the PC is temporarily unusable.