Attitude determination and disturbance estimation based on adaptive Kalman filter for nano-satellite

Nano-satellite has payload limitation to minimize weight and cost then minimum number of attitude sensors can be loaded which are sun sensor, magnetometer and gyroscope. However, The sun sensor cannot be used during eclipse, nano-satellite keeps estimating attitude with model propagator. Which contains the uncertainty, torque disturbance. For small satellite has too low moment of inertia to vulnerable to disturbance relatively. In this reason torque disturbance can be critical point to estimate attitude using propagator. The recursive attitude determination algorithm using Kalman filter is very widely used. For this reasons, attitude determination algorithms based on adaptive extended Kalman filter with disturbance torque are introduced. Which are P-scaling, Q-scaling and Q-estimation technic. In this paper, it is assumed disturbance torque represents the model uncertainty. This research shows it is possible to estimate the disturbance, angular rate and attitude of the satellites through the simulation. To solve the attitude determination problem for nano-satellite, It is possible to estimate not only attitude but also disturbance through the adaptive algorithms.