NUTS - NTNU Test Satellite, a Norwegian CubeSat Project

ADCS - Attitude determination and control

An attitude determination and control system is important for orientation control of a satellite.
An estimation method is needed for determination of the current attitude, and deviations from the expected attitude are corrected by the attitude control system.

The payload of the satellite is a camera, which requires an accurate and reliable ADC system with a pointing accuracy of less than 10 degrees in every axis. We aim to complete these control goals by using magnetorquers as actuators and a combination of a gyroscope, a magnetometer and the solar panels as sun sensors for attitude determination.

Attitude determination

Knowledge about the current attitude is required in order to control the attitude to a known reference. The chosen estimation method for the satellite is based on the QUEST (quaternion estimator) method. The method has been extended to include non-vectorized gyroscope measurements and linear prediction terms for the attitude estimation. The estimation method, EQUEST, has been compared with the well known Extended Kalman Filter as well as an xsens reference motion tracking system, and found to give accurate and reliable attitude information.


For this satellite, the solar panels already present will be utilized as sun sensors. The values for power delivered from each of the panels will be provided by other parts of the satellite and is integrated into the EQUEST algorithm. The sun vector is found via the intensity measured at each solar panel. Gyroscopes and magnetometers will be used to measure the angular velocity and the local magnetic field.

Attitude Control

Detumbling: A dissipative controller has been explored and implemented. The preliminary simulation results are satisfactory. The choice for this controller is however still left open, as use of the more popular B-dot controller might be considered a safer choice for cubesat with many other possibilities of failure.

Stabilization: The choice of a stabilization controller is very much considered a work in progress, but for the purposes of prototype testing, a reference controller has been implemented. Further work will include simulation, testing and comparing of a linear and nonlinear reference controller.