CKF 기반 상태 확장을 통해 IMU 의 장착 비정렬을 보상하는 차량용 복합항법시스템

Abstract

The navigation information of autonomous vehicles is very important not only for determining the driving route but also for controlling the vehicle. Therefore, the navigation information must exhibit a high short-term/long-term stability and accuracy, and high output frequency. Accordingly, navigation systems are configured by integrating various sensors, and a representative inertial navigation system (INS)/global positioning system (GPS) integrated navigation system has been used. In tunnels and urban areas, however, the continuity of INS error correction is not guaranteed owing to the blocking and contamination of GPS signals, and the operation of a standalone INS increases navigation error. To address this problem, the construction of an INS/GPS/non-holonomic constraints (NHC) integrated navigation based on the NHC-based movement of the land vehicle has been utilized (that is, the information in which the lateral and vertical velocities of the vehicle are 0 was used). However, the misalignment of the sensing axis of IMU with the body coordinate system of the vehicle results in the collapse of the NHC condition, which increases the INS error. To address this problem, in this study, a navigation filter was designed to estimate/compensate not only the INS errors but also the mounting misalignment of the IMU by extending the state based on the cubature Kalman filter (CKF). In addition, the accuracy of navigation information according to whether NHC information was used or not and whether it is mixed with GPS measurements was verified based on Monte Carlo simulation.