All 2015 Content
Session
Technical Session VIII: Student Competition
Abstract
Attitude determination and control systems (ADACS) are responsible for establishing desired satellite orientations. Proper satellite orientation is necessary for many science instruments and communication systems. Popular sensors include magnetometers, sun sensors, and rate gyros, and popular actuators include reaction wheels and magnetorquers. This paper investigates an ADACS design using these sensors and actuators that could feasibly be implemented on a CubeSat. The B-Dot law is used for satellite de-tumbling, and a linear inverse dynamics PD controller is utilized for steady state pointing, allowing for the analytical estimation of optimal controller gains. The inverse dynamics controller calculates desired satellite angular accelerations and then calculates the torques required to achieve these angular accelerations. This makes controller performance independent of initial conditions or system inertia properties. This system uses the magnetorquers to dump reaction wheel momentum and analyzes the satellite’s kinematic response to applied torques in order to calibrate the rate gyros and estimate system moments of inertia. Simulation results corresponded well to the analytical predictions. Often, an oscillating equilibrium would occur when controller gains were low, but this oscillation could be mitigated by selecting large controller gains such that the system was heavily overdamped and scaling down large commanded angular acceleration values to within system capabilities.
An Inverse Dynamics Satellite Attitude Determination and Control System with Autonomous Calibration
Attitude determination and control systems (ADACS) are responsible for establishing desired satellite orientations. Proper satellite orientation is necessary for many science instruments and communication systems. Popular sensors include magnetometers, sun sensors, and rate gyros, and popular actuators include reaction wheels and magnetorquers. This paper investigates an ADACS design using these sensors and actuators that could feasibly be implemented on a CubeSat. The B-Dot law is used for satellite de-tumbling, and a linear inverse dynamics PD controller is utilized for steady state pointing, allowing for the analytical estimation of optimal controller gains. The inverse dynamics controller calculates desired satellite angular accelerations and then calculates the torques required to achieve these angular accelerations. This makes controller performance independent of initial conditions or system inertia properties. This system uses the magnetorquers to dump reaction wheel momentum and analyzes the satellite’s kinematic response to applied torques in order to calibrate the rate gyros and estimate system moments of inertia. Simulation results corresponded well to the analytical predictions. Often, an oscillating equilibrium would occur when controller gains were low, but this oscillation could be mitigated by selecting large controller gains such that the system was heavily overdamped and scaling down large commanded angular acceleration values to within system capabilities.
