Cascade Attitude Observer for the SLAM filtering problem

Abstract

This article presents an attitude observer that exploits both bearing and range measurements from landmarks, in addition to reference vectors such as magnetometer and accelerometer. It is a gyro bias observer in cascade with a simplified complementary filter, driven by a gyro measurement, in which the gyro bias is estimated by comparing the bearing dynamics with the gyro measurements. The observer is compared to a full complimentary filter, and it is shown that it is more robust to initial gyro bias estimation error compared to the complimentary filter. The article also reveals how this new observer handles magnetometer failure and can use landmarks as reference vectors.