Cubesat dynamics and attitude control : Kane’s method, LQR, and Kalman filtering

Abstract

This goal of this thesis is to bring together several relevant techniques and methods commonly used in Spacecraft Attitude Control Engineering for the purpose of designing and simulating a complete CubeSat Attitude Control System. The CubeSat in question includes three reaction wheels and a rotate-able solar array (all modeled), totaling 7 degrees of freedom. The complete equations of motion are derived using Kane's method, creating a mathematical model that can be simulated and controlled. Process noise and sensor noise are then added to the system, and their effects are mitigated through the use of a multiplicative extended Kalman Filter (MEKF). Finally, multiple control methods are explored, with an emphasis on Linear Quadratic Regulation (LQR) for fine-pointing steady-state accuracy. The complete controller is then put to the test to see how it fares in three realistic mission phases: initial de-tumble, solar array pointing, and long-term disturbance rejection. Success is demonstrated in all three mission phases.