LAUSR

Abstract High-accuracy, autonomous and reliable navigation systems are important foundations for Mars rovers to achieve exploration missions successfully. Based on the motion characteristics and working environment of the rover, the paper shows a high-accuracy strapdown inertial navigation system/visual navigation system/celestial navigation system (SINS/VNS/CNS) integrated navigation scheme suitable for the rover with long-time and long-distance motion. According to the feature point positions in the camera frame at adjacent time obtained by the binocular visual odometry, its velocity in the camera frame and the attitude are calculated, then a subsystem model of SINS/VNS integrated navigation is established. In addition, using the star vectors measured by the large field-of-view star sensor, a high-accuracy attitude matrix of the rover in the inertial frame can be obtained, and a SINS/CNS subsystem model is established. Furthermore, in order to make full use of the complementary advantages of the two subsystems in attitude and position estimation, an interacting multiple model filter is developed. By updating the model probabilities of the subsystems separately in real time, the filter can output accurate navigation information of the rover. Simulation results show that the proposed navigation scheme can significantly improve the estimation accuracy of attitude, position and velocity simultaneously.