Kalman filter for attitude determination of a CubeSat using low-cost sensors

Nanosatellites have become a common choice for certain space missions, principally in universities, due to their low production and launching costs and reduced developing time. However, they present some technological challenges. In a CubeSat, total mass and available electric power are very restricted. Therefore, subsystem components and payload must be designed to have the smallest size, mass, and power consumption possible for the mission. Regarding to Attitude Determination and Control System, one of restrictions in developing CubeSats is the need to use small and low-cost sensors which fulfill satellite building restrictions. Thus, commercial-off-the-shelf components are a common choice in CubeSat development, but these parts normally present high noise levels. Therefore, a multiplicative Kalman filter for attitude determination algorithm based on quaternions is proposed. Data are simulated using typical values for low-cost sensors, namely, 3-axes magnetometer, 3-axes gyroscope, and a sun sensor. Results are compared to the extended Kalman filter proposed for AAUSAT-3 satellite and to the QUEST method, showing a better precision and low execution time compared to other methods.