Closed-Loop Attitude Determination and Control System Design of Sub-Arsec Pointing Spacecraft

ABSTRACT Spacecraft pointing accuracies with sub-arcsec to milli-arcsec levels are becoming a norm for the future space missions, mainly due to the ever-increasing demand of high resolution imaging of various celestial objects and phenomena demanded by the scientific goals of the mission. For example, imaging astronomical objects like stars including our Sun, habitable planets, nebula etc.; astrophysical phenomenon like planet, star and galaxy formations, studying the origins of life by searching for light from the first stars and galaxies after Big Bang etc.; Earth observations; imaging solar system gas planets like Jupiter, Saturn etc.; and cosmological phenomenon like dark matter etc. in various frequencies of the electromagnetic spectrum. Such tight spacecraft pointing is achieved by two methods. First, by employing mission-specific fine guidance sensor (FGS) which is the preferred method that has been utilized on most missions demanding such stringent pointing. Second is by employing only precise commercial-off-the-shelf (COTS) sensors (mainly star trackers (STR)) which is not the preferred way due to the lack of available high-precision COTS sensors though European Space Agency’s (ESA) Herschel is the only spacecraft to date that has achieved an accuracy between 0.8 and 0.9 arcsec (1s) without employing FGS and by using only precise COTS sensors. FGS is typically an interferometric instrument with complex construction (as compared to a high precision COTS STR), since it constitutes extended optics that focuses the signal from the telescope main dish to provide highly ac- curate pointing knowledge (of the order of milli-arcsec or better) of the telescope boresight error to the spacecraft for control. Construction, maintenance, in-orbit operation and calibration of the FGS is complex, time consuming and expensive as compared to the available COTS STR. This dissertation leverages the improvement and maturation of the attitude sensors (i.e. STR), the inertial sensor (i.e. gyroscope) and the reaction wheel technologies over the past 15-20 years to develop a closed-loop attitude determination and control system (ADCS) of a sub-arcsecond pointing spacecraft which can achieve the best-case pointing accuracy with the currently available high-precision COTS sensors and actuators in the American and European markets. Hence, achieving sub-arcsec attitude accuracy with already validated, reliable and high-precision COTS sensors and actuators is a step towards high-precision, modular and cheaper attitude systems for satellites in general. Moreover, the heritage ADCS algorithms employed onboard previous high-accuracy missions are utilized and improved to minimally impact already existing reliable ADCS architecture utilized by National Aeronautics and Space Administration (NASA) and ESA on various missions. The closed-loop ADCS of a spacecraft is composed of the attitude estimation system (AES); attitude control system (ACS); and external and internal disturbances acting on the spacecraft which affects its pointing accuracy. In this dissertation, AES development involves the fusion of the selected suitable and reliable high-precision star trackers and rate-integrating gyroscopes via a multiplicative extended Kalman filter (MEKF). ACS development involves the selection of the suitable and reliable COTS reaction wheels; selection of the best performing feed- back attitude controller by comparing the performances of the selected five controllers, namely, modified sliding mode control (SMC), modified linear quadratic regulator (LQR) and three modified proportional-derivative (PD) controllers; utilizing a pseudoinverse allocation algorithm for allocating spacecraft 3-axes angular momenta among three redundant reaction wheel assembly (RWA) utilized on previous missions, namely, 4-wheel pyramid RWA, 4-wheel National Aeronautics and Space Administration (NASA) skewed RWA and 6-wheel pyramid RWA configurations; and incorporating the external disturbances of the spacecraft in the ACS simulations. A complete ACS is developed by feeding the output of the best performing feedback controller to the pseudoinverse allocation algorithm to allocate the spacecraft 3-axes angular momenta among above mentioned redundant RWA. The closed-loop ADCS of the sub-arcsec pointing spacecraft is achieved by closing the loop between the AES and ACS and developing a worst-case end-to-end simulation. Actual performance characteristics of the selected hardware are incorporated in the closed-loop ADCS simulations to develop high-fidelity physically accurate simulations whose pointing performance is an indication of the best- case actual pointing that can be achieved from this closed-loop ADCS onboard a spacecraft. Two types of performance analysis, i.e. the worst-case analysis and the monte-carlo (MC) simulations analysis are carried out for this closed-loop ADCS. MC simulations of the worst-case demonstrate that an unprecedented stringent maximum three-axes absolute pointing error (APE) of [0.193, 0.078, 0.078] arcsec (3s) is achieved. This is an improvement of at least [1100, 2900, 2900]% as com- pared to ESA’s Herschel (0.8-0.9 arcsec (1s)) mission. This pointing accuracy in y- and z- axes is at least 250% better than that required by the upcoming NASA’s James Webb space telescope (JWST) mission which is scheduled to launch in 2021. Hence, JWST’s APE requirement of 0.3 arcsec (3s) can easily be achieved without using expensive and complex FGS thus saving lot of resources. These pointing accuracies achieved by the closed-loop ADCS is the best pointing accu- racy using only currently available COTS (and without using FGS) and slightly improved heritage algorithms, of all the civilian spacecraft to date (to the best of the author’s knowledge). Finally, the ADCS developed in this dissertation can be used for any other rigid spacecraft after modifying mission specific external and internal disturbances.