Abstract

The interest in small satellites for scientific missions and Earth observations has been increasing steadily in recent years and magnetic torquers have been found attractive as suitable choice of actuators for the purpose of attitude control. Magnetic torquers are commonly used for momentum desaturation of reaction wheels, damping augmentation in gravity gradient stabilized spacecraft. and reorientation of the spin axis in spin-stabilized spacecraft. Furthermore, their use as sole actuators for 3-axis stabilization of satellites in Low-Earth Orbit (LEO) has also been proven effective and advantageous when compared to other types of actuators. The autonomy of complex dynamical systems that are vulnerable to failures has been an important topic of research during the past few years. Particularly, in aerospace applications, where several constraints such as telemetry and hardware redundancy limitations make the management of on-board problems, a difficult task for ground control. With this in mind, an autonomous recovery from faults in magnetic torquers in LEO satellites constitutes the main focus of the work investigated in this dissertation. A self-recovery mechanism, which extends the capabilities of the attitude control subsystem to operate under the presence of actuator faults is developed. The solution generated takes into account the management of the control authority in the system by taking advantage of the non-faulty actuators. In other words, the recovery mechanism that is proposed in this thesis does not utilize hardware redundancy as the existing actuators are used to perform the required control action. The effects of the delay in initiating the recovery solution, the presence of noise in the magnetic field measurement, and the responses of the system that is recovered from concurrent faults are also investigated through numerical simulations. These simulations are carried out by using a model that includes relevant environmental disturbances and a realistic model of the geomagnetic field. A reduction in the average steady state error is obtained in response to and due to the application of the proposed recovery mechanism, which is applicable to the system even in the presence of fault detection delays, presence of noise in the magnetic field measurement and concurrent faults