Common Frame Based Unscented Quaternion Estimator for Inertial-Integrated Navigation

This paper explores a modified inertial-integrated navigation approach for the strapdown inertial navigation system and Global Positioning System (SINS/GPS). Since every realization of state error must be expressed with respect to the same coordinate basis, how to express the navigation parameter errors is important, especially for inertial measurement unit (IMU) errors. To this respect, the IMU errors including gyro drift and accelerometer bias expressed in common frame are more reasonable than the unframed one. This paper derives the IMU errors defined with respect to the common frame, thereby extending to propose a common frame based unscented quaternion estimator for the inertial-integrated navigation. The simulation test and car-mounted experiments for a low-cost SINS are shown to assess the performance of the new filter compared with that of the traditional filter.