Attitude and angular velocity estimation using a Luenberger observer for nonlinear systems
Nanosatellite, Attitude Determination, Nonlinear State Observer, Kalman filter.
The technological advances of the last decades, together with the standardization of the CubeSats, have made viable the use of nanosatellites for commercial and institutional purposes. Taking advantage of this trend, the National Institute of Space Research is developing the CONASAT project, which consists of a constellation of nanossatellites for the collection of environmental data. Constellation interaction requires reasonable accuracy of the satellite attitude determination and control system (ADCS), despite the limitations of the CubeSat platform and cost of the project. In this context, the present work proposes a less complex alternative to the extended Kalman filter (EKF), with an implementation of the Luenberger state observer for nonlinear systems in order to estimate the attitude and angular velocity of the satellite, considering the use of the QUEST algorithm to obtain attitude from measured vectors. The estimation scheme adopts a kinematic model of the system with parameterization of the rotation matrix by quaternion, and a dynamic model considering the object as a rigid body, with the presence of torque due to magnetic disturbances. The attitude sensor system consists of a magnetometer and a sun sensor, simulated from the ideal model with additional white noise. The implemented estimator is evaluated using simulations, together with an approach from the EKF. Finally, some conclusions and final considerations are presented.