Three-Axis Attitude Estimation Using Rate-Integrating Gyroscopes

Abstract

Traditionally, attitude estimation has been performed using a combination of external attitude sensors and internal three-axis gyroscopes. There are many studies of three-axis attitude estimation using gyroscopes that read angular rates. Rate-integrating gyroscopes measure integrated rates or angular displacements, but three-axis attitude estimation using these types of gyroscopes has not been as fully investigated. This paper derives a Kalman filtering framework for attitude estimation using attitude sensors coupled with rate-integrating gyroscopes. To account for correlations introduced by using these gyroscopes, the state vector must be augmented, compared with filters using traditional gyroscopes that read angular rates. Two filters are derived in this paper. The first uses an augmented state-vector form that estimates attitude, gyroscope biases, and gyroscope angular displacements. The second ignores correlations, leading to a filter that estimates attitude and gyroscope biases only. Simulation comparisons are shown for both filters. The work presented in this paper focuses only on attitude estimation using rate-integrating gyroscopes, but it can easily be extended to other applications such as inertial navigation, which estimates attitude and position.