Nonlinear Estimation Algorithms for two CubeSat system

Technological progress changed the satellites’ design over time: the modern tendency is to realize miniaturized spacecraft, best known as CubeSats, equipped by high-tech instrumentation and capable to perform transportation and objects’ manipulation functions, as Unmanned Aerial Vehicles or similar on the Earth. Because of their small size, these systems are subject to changes during all phases of the space mission, due to strong interaction with the external environment, internal torques and forces/torques generated by thrusters. The Attitude Determination and Control System (ADCS) has to be designed for reaching the desired state, exploiting information about inertia tensor elements, total mass and system’s centre of mass position. The purpose of this thesis is to estimate the mass properties of a rigid system, in which a collaborative satellite (chaser) and a non-cooperative one (target) are included, using nonlinear observer algorithms, for increasing the accuracy of the controller. In particular, after the definition of the system dynamics, a simple orbital simulator is designed and two different estimation algorithms are performed. First, a Recursive Least Square (RLS) Method is implemented after an appropriate manipulation of the equations of motion. Then, an Adaptive Law, based on the same rearrangement of the plant, is derived and discussed in detail. Finally, to demonstrate the effectiveness of the estimation processes, graphical and numerical results are shown after an appropriate mathematical analysis of the two defined models.