The aim of this paper is to present the contents and major findings of on-board calibration filters developed in the framework of the ESA study “Future AOCS Enabling Technologies – FEAT”. The idea behind the proposed methodology is to design filters that, while performing the canonical gyro-stellar estimation of attitude and angular rates, can estimate also attitude sensors and gyros parameters that are directly calibrated in flight. The attitude system configuration of this study is characterized by attitude sensors with at least one Star Tracker (STR) and a Fine Guidance Sensor (FGS), and one FOG redundant IMU with 4 gyros. The FGS is more accurate with respect to a STT (of the order of milli-arcsec), but it can only operate in a quasi-inertial attitude at very low angular velocity (< 1 arcsec/s) and low rate (0.5 Hz). A calibration procedure is proposed in order to estimate the mounting misalignments of STRs and gyro sensing axes, and the biases and scale factors of each gyro channels. Sensor devices introduce delays due to the acquisition and processing of the measurements. As result, the classical estimation approaches need to be re-formulated to include those delays through a fusion of the measurements. In the present study, the assumption is to have non-synchronous measurements, which leads to model the estimator managing known delays among the available sensors. In order to cope with non-synchronous measurements, the filter is designed by introducing extended state and outputs, which incorporate the delays. This work shows the design of a multi-rate EKF calibration filter (MECF) based on the range-null approach able to manage the redundancy and known delays of the avionics. A sensitivity analysis is carried out in order to establish how the filter performances change as a function of covariance values. The results show that gyros angle random walk noise and the STR noise mainly affect the filter steady-state performance. In addition, a suitable analysis is conducted to ensure the dynamic observability of MECF.
ON-BOARD REAL-TIME CALIBRATION OF NON-SYNCHRONOUS ATTITUDE SENSORS AND GYROS, / Curti, Fabio; Toglia, Chiara.; Pererez Gonzalez, Jose Alvaro; Sechi, Gianfranco. - (2021). (Intervento presentato al convegno ESA GNC 2021 Conference tenutosi a Virtual).
ON-BOARD REAL-TIME CALIBRATION OF NON-SYNCHRONOUS ATTITUDE SENSORS AND GYROS,
Curti, Fabio
Primo
Conceptualization
;
2021
Abstract
The aim of this paper is to present the contents and major findings of on-board calibration filters developed in the framework of the ESA study “Future AOCS Enabling Technologies – FEAT”. The idea behind the proposed methodology is to design filters that, while performing the canonical gyro-stellar estimation of attitude and angular rates, can estimate also attitude sensors and gyros parameters that are directly calibrated in flight. The attitude system configuration of this study is characterized by attitude sensors with at least one Star Tracker (STR) and a Fine Guidance Sensor (FGS), and one FOG redundant IMU with 4 gyros. The FGS is more accurate with respect to a STT (of the order of milli-arcsec), but it can only operate in a quasi-inertial attitude at very low angular velocity (< 1 arcsec/s) and low rate (0.5 Hz). A calibration procedure is proposed in order to estimate the mounting misalignments of STRs and gyro sensing axes, and the biases and scale factors of each gyro channels. Sensor devices introduce delays due to the acquisition and processing of the measurements. As result, the classical estimation approaches need to be re-formulated to include those delays through a fusion of the measurements. In the present study, the assumption is to have non-synchronous measurements, which leads to model the estimator managing known delays among the available sensors. In order to cope with non-synchronous measurements, the filter is designed by introducing extended state and outputs, which incorporate the delays. This work shows the design of a multi-rate EKF calibration filter (MECF) based on the range-null approach able to manage the redundancy and known delays of the avionics. A sensitivity analysis is carried out in order to establish how the filter performances change as a function of covariance values. The results show that gyros angle random walk noise and the STR noise mainly affect the filter steady-state performance. In addition, a suitable analysis is conducted to ensure the dynamic observability of MECF.I documenti in IRIS sono protetti da copyright e tutti i diritti sono riservati, salvo diversa indicazione.
