An estimator for Attitude and Heading Reference Systems based on Virtual Horizontal Reference

Abstract

The output of the attitude determination systems suffers from large errors in case of accelerometer malfunctions. In this paper, an attitude estimator, based on Virtual Horizontal Reference (VHR), is designed for an Attitude Heading and Reference System (AHRS) to cope with this problem. The VHR makes it possible to correct the output of roll and pitch of the attitude estimator in the situations without accelerometer measurements, which cannot be achieved by the conventional nonlinear attitude estimator. The performance of VHR is tested both in simulation and hardware environment to validate their estimation performance. Moreover, the hardware test results are compared with that of a high-precision commercial AHRS to verify the estimation results. The implemented algorithm has shown high accuracy of attitude estimation that makes the system suitable for many applications.